Hypogea 2023

50 Artificial cavities and geo-risk assessment: the case of “The Strade Nuove and the system of the Palazzi dei Rolli” With an original core of 113 hectares, Genoa’s historical centre is considered as one of the largest ancient city centres in Europe. In addition, the high density of buildings, especially after a significant urban growth process that began in the 18th century, makes it one of the most densely populated historical centres: about 23,000 inhabitants live in the oldest area (including the port), distributed in 2305 buildings (value as of 1999) on a volume of about 10 million cubic meters. The area is crossed by several streams, completely covered since historical times, whose presence is sometimes suggested by toponymy (Bixio et al., 2015): from east to west, we can describe the Torbido, the Sant’Anna, the Carbonara (which also includes the San Gerolamo), the Sant’Ugo, and the Rio Lagaccio stream, which flows along the edge of Palazzo del Principe (Bixio et al., 2017). The geological setting is relatively simple, as it is represented by two rock masses that are completely different in composition and susceptibility to erosion (fig. 3): a flysch formation, made up of marly limestones and limestone marls with shales interlayers, arenites, and marls, constitutes a hard rock mass of fair geo-mechanical quality and it dominates the sectors at higher elevations such as the Carignano hill in the south or the Castelletto-Principe hill in the northwestern sector; the argille di Ortovero, represented by over-consolidated fissured clays with a geotechnical behaviour intermediate between a rock and a soil, show significantly gentler morphologies and characterise the historical centre between Largo della Zecca- Piazza Fontane Marose, Porta Soprana and Piazza Cavour (Faccini et al., 2021). The contact between the two formations, with strong hydrogeological contrast, is often marked by historical wells and springs. The UNESCO Site “The Strade Nuove and the system of the Palazzi dei Rolli” extends across sections Fig. 3 – Geomorphological sketch map. 1) man-made landforms (fills, sea embankments, dumps); 2) stiff fissured clays; 3) marly limestone with shales interlayers, siltstone, marls (graphics A. Ferrando).

51 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa of the historic centre of Genoa, and it represents the first European example of an urban development project with a unified structure, planned by a public authority and associated with a particular system of public hospitality in private residences (fig. 4). The Site includes a collection of Renaissance and Baroque palaces along the so-called “Strade Nuove”, which represent a hinge between the medieval streets to the south and contemporary traffic structures to the north: the “Strade dei Rolli” are Fig. 4 – Le Strade Nuove and the system of the Palazzi dei Rolli World Heritage area (graphics A. Ferrando): 1) Palazzo di Antonio Doria; 2) Palazzo di Clemente della Rovere; 3) Palazzo di Giorgio Spinola; 4) Palazzo di Tomaso Spinola; 5) Palazzo di Giacomo Spinola; 6) Palazzo di Agostino Ayrolo; 7) Palazzo Paolo e Niccolò Interiano; 8) Palazzo Agostino Pallavicino; 9) Palazzo di Pantaleo Spinola; 10) Palazzo di Franco Lercari; 11) Palazzo di Tobia Pallavicino; 12) Palazzo Angelo Giovanni Spinola; 13) Palazzo di Gio Battista Spinola; 14) Palazzo di Nicolosio Lomellino; 15) Palazzo di Lazzaro e Giacomo Spinola; 16) Palazzo di Niccolò Grimaldi; 17) Palazzo di Baldassarre Lomellini; 18) Palazzo di Luca Grimaldi; 19) Palazzo di Francesco e Ridolfo Brignole Sale; 20) Palazzo di Gerolamo Grimaldi; 21) Palazzo di Gio. Carlo Brignole; 22) Palazzo di Bartolomeo Lomellino; 23) Palazzo di Stefano Lomellini; 24) Palazzo di Giacomo Patrone Lomellini; 25) Palazzo di Antoniotto Cattaneo; 26) Palazzo di Gio. Agostino Balbi; 27) Palazzo di Gio. Francesco Balbi; 28) Palazzo di Giacomo e Panaleo Balbi; 29) Palazzo di Francesco Maria Balbi Piovera; 30) Palazzo di Stefano Balbi; 31) Palazzo di Cosmo Centurione; 32) Palazzo di Giorgio Centurione; 33) Palazzo di Gio Battista Centurione; 34) Palazzo di Cipriano Pallavicini; 35) Palazzo di Nicolò Spinola; 36) Palazzo di Francesco Grimaldi; 37) Palazzo di Gio Battista Grimaldi; 38) Palazzo di Gio. Battista Grimaldi; 39) Palazzo di Stefano De Mari; 40) Palazzo di Ambrogio De Nigro; 41) Palazzo di Emanuele Filiberto Di Negro; 42) Palazzo di Croce De Marini.

52 Artificial cavities and geo-risk assessment: the case of “The Strade Nuove and the system of the Palazzi dei Rolli” Fig. 5 – Images of hypogea of the “Strade Nuove and system of the Palazzi dei Rolli World Heritage” area: A) one of the entrances of the bunker of Villetta Di Negro (photo R. Bixio); B) cistern of Palazzo Gio. Battista Grimaldi [n. 38 in Fig. 3] (photo M. Traverso); C) putridarium under the church of San Filippo Neri (photo R. Bixio); D) cistern of former convent of San Silvestro (photo S. Saj); E) large underground pond of spring water under Palazzo Paolo e Niccolò Interiano [n. 7 in Fig. 3] (photo A. Figari).

53 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Strada Nuova (Via Garibaldi), Via San Luca/Piazza De Marini, Salita Santa Caterina/Piazza Fontane Marose, Via Balbi, Via Lomellini/Piazza Fossatello, Largo della Zecca/Piazza della Nunziata, Strada Nuovissima (Via Cairoli). The Palazzi dei Rolli were built as residences of the wealthiest and most powerful aristocratic families of the Republic of Genoa and were ranked based on the value of the land and the quality of the building in order to be entered on lists or registers, called “Rolli”, for public hospitality. The inclusion of the Site “The Strade Nuove and the system of the Palazzi dei Rolli” in the UNESCO World Heritage List took place in 2006. There are 42 Palazzi dei Rolli, and the streets with the greatest concentration of noble residences are Via Garibaldi, Via Balbi, Via Lomellini and Salita Santa Caterina. Many buildings have preserved the original structure almost intact and can be visited in part as they are home to public institutions and museums. These include those belonging to the Strada Nuova Museum system-Palazzo Rosso, Palazzo Bianco and Palazzo Tursi-Palazzo Spinola di Pellicceria, home of the National Gallery of Palazzo Spinola, and Palazzo Reale. Artificial cavities in the Genoa historical centre The survey and inventory of artificial cavities in Genoa’s historical centre is an activity that have been undertaken more than thirty years ago by some of the founding members of Centro Studi Sotterranei di Genova (Melli et al., 2006). A particular boost to the survey was provided with the CIVIS project of the Municipality of Genoa (2000), which enabled the first systematic survey of the culverted watercourses that characterise the morphological amphitheatre on which the area of Genoa’s historical centre has been settled over time and that undoubtedly represent the most extensive type of artificial cavity in the territory surveyed (Bixio et al., 2015). Restricting the analysis to the area of “The Strade Nuove and the System of the Palazzi dei Rolli” that are part of the UNESCO World Heritage List and referring to the typology of artificial cavities proposed by the Italian Speleological Society (Parise et al., 2013) the census collected up to this day consists of underground voids attributable to hydraulic underground works, hypogean civilian dwellings, religious work, military and war works and transit (fig. 6). Fig. 6 – Artificial cavities sketch map in the Genoa old city (graphics A. Ferrando).

54 Artificial cavities and geo-risk assessment: the case of “The Strade Nuove and the system of the Palazzi dei Rolli” Among the typology related to hydraulic works, 14 drainages, 8 sewer conduits, 12 culverted watercourses, 4 wells and 49 cisterns were surveyed and registered; particularly significant is the hydrographic network, as mentioned in the previous section. The length of the culverted stretches varies between 1 and 2 km, and many streams flow under buildings of historicalmonumental interest, such as S.Ugo Stream flowing under the Commenda di Pre or Sant’Anna Stream flowing under Palazzo San Giorgio and others. Worth mentioning are wells and cisterns, the current inventory of which is most likely significantly underestimated. In fact, it must be assumed that many buildings in the historic centre either have an underground cistern as an annex or had a well in historic times for potable use, the survey of which, however, is difficult due to access on private property. Among the most significant cisterns surveyed, a brief mention can be given to the one in Piazza Fontane Marose, in the middle of the Unesco World Heritage area: recent investigations have verified that it is most likely a well, dug in the medieval period exactly at the contact between marly limestones and over-consolidated clays, whose hydrogeological significance was mentioned earlier. Among man-made cavities for civilian use, at least 44 shelter tunnels have been listed and surveyed, all of which were built between 1941 and 1943 by different institutions, including the City of Genoa, the German TODT Organization, and the Fire Brigade, which designed and excavated many tunnels with the function of air-raid shelters (Bixio et al., 2019). Their area covers more than 100,000 m2 spread over more than 30 km of underground environment that could exceptionally accommodate more than 100,000 people (Bixio et al., 2019). Some of these shelter tunnels are now repurposed as elevators and connections for public use, underground garages, and other uses. Among the category of worship-related works, 8 crypts were surveyed, generally of limited size (within 50 m2 ), some of them of great architectural value and connected with multiple pits/tombs with interspaces. It is likely that even in this case the number of voids recognized is significantly lower than the actual number, given the large number of churches and places of worship built in Genoa’s historic centre since historic times. Among the underground works for military use, 18 shelter tunnels for military use have been recorded and surveyed, with development and dimensions similar to those described for civilian use (Bixio et al., 2011; 2012), while among transit works, 14 tunnels have been surveyed, some of which are now used for ordinary roads. This preliminary database will provide insight into specific urban planning interventions aimed at the conservation and management of Genoa’s artificial cavities, which represent a unique cultural heritage and a resource for their historical and socio-economic importance. First georisk assessment and final remarks The preliminary inventory of artificial cavities in Genoa’s historical centre has identified more than 150 artificial cavities, many of them with significant linear development, such as the case of the hydrographic network or the anti-aircraft shelter tunnels. Many of these voids are to this day poorly known and, in some cases, unknown, as is the case of Palazzo Belimbau (or Palazzo Cattaneo, n. 25 in fig. 3), which dates back 1594. The building has a cistern of 200 m3 in size (the entrance to which opens behind a marble slab on the ground floor) and a shelter-tunnel that Fig. 7 – Images of hypogea of the “Strade Nuove and system of the Palazzi dei Rolli World Heritage” area: A) former Romanesque basilica formerly re-used as final stretch of the eastern branch of Acquedotto Storico: this cistern is known amongst the Genoese as “Lampionea” (photo M. Traverso); B) entrance to the cistern of Palazzo Negrone, formerly Palazzo di Agostino Ayrolo [n. 25 of Fig. 3] (photo R. Bixio).

55 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa can be traced back to the World War II period (the entrance to which opens near the elevator shaft, also on the ground floor). All surveyed artificial cavities were mapped through GIS by introducing the actual measurements of the surveyed voids, where known, or otherwise assigning a buffer of 3 m in case of a linear element and 5 m in case of a point element. The resulting database was then crossed with the layer related to buildings in Genoa’s historical centre, highlighting those located above underground voids potentially at risk of sinkhole phenomena and further distinguished between Palazzi dei Rolli and the Strade Nuove and other buildings in the historic centre otherwise subject to historical-monumental protection (fig. 8). A preliminary analysis has revealed that some sections of “Strade Nuove” included in the UNESCO World Heritage Site, such as Strada Nuova (Via Garibaldi), Via San Luca and Piazza De Marini, Salita Santa Caterina and Piazza Fontane Marose, Via Balbi, Via Lomellini and Piazza Fossatello, Largo della Zecca and Piazza della Nunziata lie on underground voids: these are mainly wells, cisterns, war tunnels and culverted waterways. From this analysis it emerges that numerous buildings included in the Palazzi dei Rolli are characterised by underlying voids: in the section of Salita Santa Caterina and Piazza Fontane Marose there are Palazzo Negrone (Ayrolo Agostino), Palazzo Interiano-Pallavicini and Palazzo Spinola di Luccoli Tedeschi. In StraFig. 8 – Geo-risk sketch map (graphics A. Ferrando).

56 Artificial cavities and geo-risk assessment: the case of “The Strade Nuove and the system of the Palazzi dei Rolli” da Nuova (Via Garibaldi) they include Palazzo Cambiaso Pallavicini, Palazzo Pantaleo Spinola or Gambaro, and Palazzo Carrega-Cataldi (formerly Tobia Pallavicini). Between Via San Luca and Piazza de Marini, Palazzo Gio Battista Grimaldi, Palazzo di Stefano De Mari and Palazzo De Marini-Croce are affected by artificial cavities. Between Largo della Zecca and Piazza dell’Annunziata, Palazzo Cattaneo Belimbau, Palazzo Doria Lamba, Palazzo Rostan Reggio and Palazzo Lomellini Patrone, are built over underground voids, while other significant cases between Via Lomellini and Piazza Fossatello are Palazzo Centurione Durazzo Pallavicini, Palazzo Centurione and Palazzo Pallavicini. Lastly, in Via Balbi, Palazzo Cattaneo and Palazzo Reale (Balbi Durazzo) are characterised by underground voids. This study must be considered as an early risk assessment analysis for the historic centre of Genoa and has made it possible to highlight the need for an in-depth knowledge of the area in order to better manage the cultural and landscape heritage, also in light of the effects of climate change, which in the case under study is expressed by an increase in average air temperature and a significant change in the rainfall regime. This opens up multiple research perspectives, among which the first in order of priority is the need to complete the census and detailed survey of all artificial cavities in Genoa’s historical centre: the involvement of scientifically qualified and experienced speleologists of artificial cavities, the only category of experts able to move with autonomy and safety in underground and difficult environments, is indispensable for such in-depth investigations, in order to collect information and surveys of fundamental importance both for the assessment of the hazard from sinkhole and mitigation of the related risk and for the cultural enhancement of the hypogea. A further step could be to design and implement ground and buildings monitoring activities through new technological solutions, in situ and remote sensing, both for knowledge purposes and possibly for warning and civil protection. Bibliography Argentieri A., Occhigrossi B.C., Piro M., Rotella G., 2018, Natural and anthropogenic cavities and sinkholes in Rome metropolitan area: From geological and speleological research to land management, Rendiconti Online Societa Geologica Italiana, 44, pp. 104-111. Bixio R., Faccini F., Maifredi A., Perasso L., Saj S., Traverso M., 2017, The culverted streams in the historical amphitheatre of Genoa city (Italy): flood risk or geoheritage protection?, Hypogea2017, Proceedings of International Congress of Speleology in Artificial Cavities – Cappadocia, Turkey, March 06/10, pp. 165-176. Bixio R., Faccini F., Perasso L., Piana P., Saj S., Traverso M., 2019, Second World War Air-Raid Shelters in Genoa (Italy): knowledge, protection and use of an underground historical and cultural heritage in urban environment. Hypogea 2019, Proceedings of International Congress of Speleology in Artificial Cavities – Dobrich, May 20-25, pp. 191-196. Bixio R., Ferrando L., Saj S., Traverso M., 2006,Genova ipogea: esplorazioni nei sotterranei del Ghetto. In: Buti A. (Ed.) Convegno “il quartiere del Ghetto di Genova. Studi e proposte per il recupero dell’esistente”, pp. 113-121. Collana Architettura e Restauro, Arkos, Nardini Ed., Firenze. Bixio R., Saj S., Traverso M., 2011, Air-raid shelters of the second world war in Genoa: the bunker of Prefecture. VIII Convegno Nazionale di Speleologia in Cavità Artificiali, Ragusa, Italy, Speleologia Iblea, 15-2011/13, pp. 171-178. Bixio R., Saj S., Traverso M., 2012, Il bunker della Prefettura di Genova. La Casana, 2/2012, pp. 18-23. Bixio R., Saj S., Traverso M., 2015, Urban hydrographic network of Genoa’s historic centre: the underground course of the Fossatello stream, in: Parise M., Galeazzi C., Bixio R. & Germani C. (Eds.) - Proceedings of the International Congress in Artificial Cavities “Hypogea 2015”, pp. 129-140, Rome, March 11-17. Brandolini P., Cappadonia C., Luberti G.M., Donadio C., Stamatopoulos L., Di Maggio C., Faccini F., Stanislao C., Vergari F., Paliaga G., Agnesi V., Alevizos G., Del Monte M., 2020, Geomorphology of the Anthropocene in Mediterranean urban areas, Progress in Physical Geography, 44(4), pp. 461-494. Brandolini P., Faccini F., Paliaga G., Piana P., 2018. Man-Made landforms survey and mapping of an urban historical centre in a coastal Mediterranean environment. Geogr. Fis. Din. Quat., 41, pp. 97-102. Faccini F., Giardino M., Paliaga G., Perotti L., Brandolini P., 2021, Urban geomorphology of Genoa Old City (Italy) , Journal of Maps, 17(4), pp. 51-64. Grossi Bianchi L., Poleggi E., 1971, Una città portuale del medioevo, Genova nei secoli X-XVI, 339 pages, Sagep Editrice, Genova. Melli P., Bixio R., Saj S., Traverso M., Ferrando L., 2006, Genova sotterranea, pp. 37-79, Erga Edizioni, Genova. Parise M., Galeazzi C., Bixio R., Dixon M., 2013, Classification of Artificial Cavities: A First Contribution by the UIS Commission, in Proceedings of the 16th International Congress of Speleology, edited by M. Filippi and P. Bosak, 2, pp. 230–235, Brno: Czech Speleological Society, July 21–28. Terrone M., Piana P., Paliaga G., D’Orazi M., Faccini F., 2021, Coupling Historical Maps and LiDAR Data to Identify Man-Made Landforms in Urban Areas, ISPRS Int. J. Geo-Inf. 2021, 10, 349.

57 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Dipartimento di Scienze della Terra e Geoambientali, Università degli Studi di Bari Aldo Moro, Bari, Italy 2 CNR-IRPI, Italy 3 Dipartimento di Ingegneria, Università degli Studi della Campania “Luigi Vanvitelli”, Aversa, Italy 4   Autorità di Bacino Distrettuale dell’Appennino Meridionale, Caserta, Italy *  Reference author email: [email protected] A chronology of sinkholes related to artificial cavities in the hydrographic district of the Southern Apennines of Italy Isabella Serena Liso1 , Carmela Vennari2 , Maria Assunta Fabozzi3 , Daniela Ruberti3 , Marco Vigliotti3 , Gennaro Capasso4 , Vera Corbelli4 , Mario Parise1,* Abstract Occurrence of sinkholes related to presence of artificial cavities is a very common hazard in many towns and villages of Italy, and is not limited to karst settings. Actually, the use of underground by humans involved a large part of the Italian territory, due to its very long historical and cultural history. At some locations, such a use has been so extensive that, in the historical part of towns, it is very difficult to find areas without subterranean voids. In the recent years, a chronological catalogue of sinkholes (of both natural and anthropogenic origin) in Italy has been published, based upon the classification of artificial cavities developed by the specific Commission of the Italian Speleological Society (Galeazzi, 2013), later on adopted at the international level by the Commission on Artificial Cavities of the International Union of Speleology (Parise et al., 2013). In this contribution we intend to extract and discuss the sinkholes evolution in time, related to artificial cavities. The study area covers one of the largest hydrographic districts of Europe, that is the Hydrographic District of Southern Apennines in Italy. It includes five entire Regions (Puglia, Calabria, Campania, Basilicata, Molise) and a portion of Lazio and Abruzzo Regions. Even though the catalogue is inevitably affected by inhomogeneity, due to the variety of sources used to fill the database, and the possible incompleteness of data, it offers a first analyses on the effect that sinkholes related to artificial cavities may produce on the society and infrastructures. A good knowledge of the cavity genesis and timing of sinkhole phenomena occurred in the past is the first step to be reached in order to improve the societal resiliency toward natural and anthropogenic hazards; this is particularly true in complex geological settings such as the territory of the Italian Hydrographic District of Southern Apennines. Keywords: artificial cavities, sinkholes, damage, susceptibility, hazard. Introduction Sinkholes (Gutierrez et al., 2014; Parise, 2019, 2022) are among the main types of geological hazards directly related to the underground environment. They represent a subtle and dangerous hazard, generally developing without tangible evidence, with a final, catastrophic phase typically occurring with very little precursory signs, thus creating a high risk to the built-up areas (Waltham et al., 2005; Parise and Gunn, 2007; Parise, 2015). Countries characterized by a long history, such as Italy, where the cultural vicissitudes brought to development of many different civilizations through the millennia, are inevitably characterized by a diffuse and variegated use of the underground. Especially at those sites where the local geological conditions allowed (Palmer, 2007; Del Prete and Parise, 2013), man used to excavate soft materials as volcanic rocks, calcarenites, tufa, etc., in order to realize artificial cavities. Variety of these latter has been properly pointed out in the Classification of Artificial Cavities, initially proposed by Italian scholars (Galeazzi, 2013), and later on adopted at the international level by the specific Commission of the International Union of Speleology (UIS; Parise et al., 2013). In this classification, each category is designated with a capital letter, to which a number is added for further specification of the purpose according to which the original cavity was dug. This system easily allows to attribute each artificial cavity to a category, even though it should be considered the frequent possibility of changes in the use of the cavity throughout its history. In terms of land management and planning, what is more important, however, is the exact location of underground voids, and their present conditions. As a matter of fact, many artificial cavities, once abandoned, may soon became sites of degradation of the rock mass, suffering decaying process that may potentially bring to a decrease in the physical properties of the rocks, mostly due to water infiltration and weathering processes (Canakci, 2007; Calcaterra and Parise, 2010; Castellanza et al., 2018). In time, rock mass degradation might contribute to the progressive development of instability,

58 A chronology of sinkholes related to artificial cavities in the hydrographic district of the Southern Apennines of Italy starting underground and then progressing toward the surface, until producing a real sinkhole (Fraldi and Guarracino, 2009; Lollino et al., 2013). The Hydrographic District of Southern Apennines (from now on, ADAM, its acronym in Italian language) includes five whole Regions (Apulia, Calabria, Campania, Basilicata, Molise) of Peninsular Italy, and portions of Latium and Abruzzo. It is an inter-regional body in charge of, among many other purposes, dealing with geo-hydrological hazards and defining the related regulations for land management. Within the framework of a project dedicated to cavities in the ADAM territory, we present in this contribution some preliminary data about the spatial and chronological distribution of sinkholes linked to artificial cavities in the ADAM Hydrographic District, as a first step in the process of improving the knowledge about occurrence of this subtle hazard. In detail, while waiting for availability of further data from the single municipalities, collected at ADAM but still not fully organized as a digital database, we started our analysis by using the datasets in our possession, and mainly the chronological database on sinkholes in Italy, popoulated since several years by a joint project between the Institute of Research for Geo-Hydrological protection of the National Research Council of Italy and the Earth and Environmental Sciences Department at Aldo Moro University in Bari. This database has already been object of some publications, as concerns both the overal catalogue (Parise and Vennari, 2013; Vennari and Parise, 2022) and single regions, with particular regard on Apulia (Parise and Vennari, 2017; Vennari et al., 2022). Anthropogenic sinkholes in the ADAM territory In this section, we examine and extract from the catalogue the events of sinkholes related to artificial cavities occurred within the Hydrographic District of Southern Apennines in Italy, and discuss their evolution in time, and the consequences on society. Even though the catalogue is inevitably affected by inhomogeneity, due to the variety of sources used to populate the database (including chronicles, historical documents and books, archival research, techinal reports, scientific publications, speleological surveys, etc.), and to incompleteness of data, it offers a first analyses on the effect that sinkholes related to artificial cavities may produce on society and infrastructures. A good knowledge of the cavity genesis and timing of sinkhole phenomena occurred in the past is the first step to be reached to improve the societal resiliency toward natural and anthropogenic hazards; this is particularly true in complex geological settings such as the territory of the Italian Hydrographic District of Southern Apennines. Since quality of data plays an important role in the construction of any dataset, the chronological catalogue of sinkholes in Italy included only those events for which a temporal reference was available. This latter could be the exact time of occurrence (hour, day, month, year), but also simply be the year, or a range between years, and so on. To evaluate the level of knowledge of these parameters, the concepts of accuracy, certainty and reliability were adopted (for further details, see Vennari and Parise, 2022). In particular, accuracy was considered in reference to the time of sinkhole occurrence, the true value being represented by the complete knowledge of its time of occurrence. When all these data are available, the accuracy is high, whereas it decreases when the data move away from the full knowledge (table 1). Temporal accuracy Available information Examples High Hour-day-monthyear 8:10 am, 20 February 2020 Middle-High Day-month-year 20 February 2020 Middle Month (or season)- year February 2020, or Winter 2020 Middle-Low Year 2020 Low Range of years Between 2019 and 2020, or after 2020 Tab. 1 – Temporal accuracy ranges (after Vennari and Parise, 2022). According to the most recent update of the catalogue (June 2023), the documented sinkholes in the ADAM territory are 752, including in this count both the events related to natural karst caves and to man-made cavities (plus a small amount of unknown origin). Sinkholes affect the whole territory of ADAM, with a clear prevalence of those related to man-made cavities (about 69%), with about 14% originated by natural karst caves. For 129 events, corresponding to 17%, the origin of the documented sinkholes is unknown. Excluding the natural sinkholes, the number reduces to 664 events, which geographical distribution is listed in table 2 and shown in figure 1. Region Sinkholes related to artificial cavities Lazio 2 Abruzzo 16 Molise 3 Campania 525 Apulia 102 Basilicata 6 Calabria 10 Tab. 2 – Regional distribution of sinkholes related to artificial cavities (from scrutiny of the chronological catalogue, see text for further details).

59 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Fig. 1 – Chronological distribution of documented artificial sinkholes in the ADAM territory.

60 A chronology of sinkholes related to artificial cavities in the hydrographic district of the Southern Apennines of Italy It appears that the largest number of events regards Campania (525 sinkholes), followed by Apulia (102), whilst the other regions show a remarkably lower number of documented sinkholes. This is mostly due to the great attention paid in Campania (particularly in Naples) on sinkholes in urban areas, where a specific office was established to collect data and to face the problem. The availability of recent scientific publications on the area further allowed to have a rich documentation at Naples (Guarino and Nisio, 2012; Guarino et al., 2018; Tufano et al., 2022). As represented in the pie chart for Campania in figure 1, the province of Naples is by large the area showing the highest percentage of sinkholes (95%) in the region. In Apulia, the research interest by scholars to the issue of sinkholes, on one side, and the activity by the past Basin Authority (nowadays included in ADAM), on the other, made possible to monitor the territory at a high detail, thus collecting a significant number of events with chronological reference (Delle Rose et al., 2004; Fiore and Parise, 2013; Fiore et al., 2018). As a matter of fact, the Apulia pie chart in figure 1 shows documentation of sinkholes in all provinces, testifying a complete coverage of the territory by the phenomenon, and a detailed data collection by researchers as well. Such a work was not carried out with the same attention in other regions where, inevitably, the numbers appear lower. Such a discrepancy will hopefully be, at least partly, solved by examining and including in our dataset further data from ADAM in the following years, including those not having a chronological documentation. As for the temporal occurrence of artificial sinkholes (figure 2), the oldest event dates back to the XIV century, corresponding to a sinkhole in the Benevento province (Campania). Then, other documented anthropogenic sinkholes are reported, typically with low numbers, until 1950s. Since those years, the number definitely increases, reaching its peak in the last decades, with a frequency at least doubling with respect to the pre-2000 years. This is at least in part due to greater awareness of the problem, after a series of events that hit several localities, from Gallipoli in 2007 and Altamura since 2008 in Apulia (Parise, 2012; Pepe et al., 2013), or the city of Naples in different years (see before). In the aftermath of these (and others) events, a more careful record of the occurrence of sinkholes was produced in many regions of southern Italy, especially where specific regulations were issued by local authorities. Importance of cave surveying, through specific mapping of the features related to instability, must be highlighted in order to reach a good knowledge about the processes occurring underground (Klimchouk and Andrejchuk, 2002; Palmer, 2007), and particularly on the few observable precursory signs (Swedzicki, 2001; Parise and Lollino, 2011; Parise, 2015, 2017). Fig. 2 – Chronological distribution of documented artificial sinkholes in the ADAM territory.

61 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Among the categories of artificial cavities, underground quarries are definitely the type which has been demonstrated in many geological settings as the most responsible for sinkhole occurrence (Hutchinson et al., 2002; Ferrero et al., 2010). This occurred even in southern Italy (Parise, 2010, 2012; Pepe et al., 2013), so that this category of underground works has become the main field of application of numerical codes to model the evolution of underground instabilities, and to forecast the size of sinkholes at the ground surface (Parise and Lollino, 2011; Lollino et al., 2013; Fazio et al, 2017; Perrotti et al., 2018; Lollino and Parise, 2023). In the 664 events documented in the ADAM territory, for a small number it was possible to ascertain the category of artificial cavities at the origin of the failure. Nevertheless, the large majority of sinkholes were related to mining works, with particular regard to categories E1, E3 and E5. Secondarily, categories B4 and B9 were also involved. Origin of natural sinkhole events has to be related to different genesis processes (Gutierrez et al., 2014; Parise, 2019, 2022), that cause at the ground effects varying from mild depressions, slightly lower than the surrounding terrain, to abrupt features with steep walls and depth greater than several tens of meters. In addition to natural karst sinkholes, in the last decades the reports about anthropogenic sinkholes (related to man-made cavities) have definitely increased. Artificial cavities are extremely widespread in Italy, due to history of the country, that, combined with its geological and morphological characters, greatly favored the development of cave-dwelling civilizations (Fonseca, 1980; Laureano, 1993; Varriale et al., 2022; Bixio et al., 2023). An important aspect is related to the main sinkhole triggering factors; this information is difficoult to collect. Nevertheless, out of the 664 events, for 337 the triggering factors have been recognized (figure 3). Rainfalls represent definitely the most common trigger, since they appear to have caused about 220 events, corresponding to 65% of those for which a trigger is indicated. Among the other triggering factors, human activities related to presence of network infrastructures, direct collapses of existing cavities, maintenance works, and vibrations due to traffic are present, in decreasing percentage. The documented sinkholes had heavy negative effects on economy and society. These impacts and effects of sinkole occurrances on people are summarized in figure 4. Even in this case, due to the claer majority of consequences for the Campania events, the histogram regarding Campania is represented separately from the others, due to different scales. Discussion The data presented in this contribution represent a percentage of what actually occurred in the ADAM territory as concerns the frequency of sinkholes linked to artificial cavities. This because in this contribution, we took into account only the sinkhole events with a temporal reference (see the chronological catalogue of sinkholes of Italy; Vennari and Parise, 2022). NotFig. 3 – Main triggering factors inducing the occurrence of documented sinkholes in the ADAM territory, as extracted from the chronological database.

62 A chronology of sinkholes related to artificial cavities in the hydrographic district of the Southern Apennines of Italy withstanding this limitation, the outcomes indicate a diffuse occurrence of artificial sinkohles, and point out to the need for further studies and researches, aimed at reducing the risk from future collapse events. This ambitious goal must be reached only taking into account the difficulties in the identification of underground sites, that in many cases have been abandoned by man, and of which the memory has been lost. It is of crucial importance the link with speleologists, since they are able to move in safety underground, and have built through the years a wonderful register of natural caves and artificial cavities for the whole Italian territory. At this regard, it will be important in the next weeks to integrate our dataset with data coming from the inventory of caves by the Italian Speleological Society (SSI), thanks to an agreement recently signed between ADAM and SSI. Sharing of the data will allow to better connect the sinkholes with the presence of known cavities, or, alternatively, to perform research and exploration to identify other possible cavities and to survey them. In case a new cavity is found, in fact, the first necessary activity is to ascertain its spatial development, and to verify the possibility of failures and/or the likely involvement of nearby infrastructures. Nowadays, speleological surveys are carried out with high degree of precision, following international standards (Day 2002; Häuselmann 2006, 2010). In addition to such a link, co-operation with local archives, and research on historical sources, is another issue worth to be strengthened, given the remarkable documentation available in Italy. Historians may provide a fundamental help toward a correct research on, and interpretation of, the historical and archival sources, and come out with further data about sinkholes. Conclusions Instability occurring underground may eventually result in producing subsidence or sinkholes at the ground surface over large areas, and may be at the origin of serious damage to society, and to casualties as well. In inhabited areas with a high number of man-made cavities, this poses serious problems in terms of Civil Protection issues. Within the framework of a project funded by the Hydrographic District of the Southern Apennines, we have presented in this work a preliminary assessment concerning the chronological distribution of artificial sinkholes in the ADAM territory. It is a first step in the research about sinkholes, which will be followed by further analysis, dedicated to geology of the territory and definition of procedures for the preliminary evaluation of the artificial cavity’s stability conditions; detailed assessment at two pilot areas within the ADAM territory, including modelling analyses, and establishment of actions addressed toward mitigation of the risk related to sinkholes. Fig. 4 – Societal impacts from artificial sinkholes in the ADAM territory. Due to the high number of data (and the different scale), Campania region is shown separately.

63 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Acknowledgments This work was partly carried out within the framework of two joint contracts between the Hydrographic District of Southern Apennines and, respectively, the Earth and Environmental Sciences Department at University Aldo Moro in Bari (scientific responsible: M. Parise) and the Engineering Department at University “Luigi Vanvitelli” in Aversa (scientific responsible: D. Ruberti). References Bixio R., De Pascale A., Galeazzi C., Parise M., 2023, Rupestrian works and artificial cavities: categories of construction techniques. Journal of Architectural and Engineering Research, 4, 58-69. Calcaterra D., Parise M. (Eds.), 2010, Weathering as a predisposing factor to slope movements. Geological Society of London, Engineering Geology Special Publication no. 23. Canakci H., 2007, Collapse of caves at shallow depth in Gaziantep City center, Turkey: a case study. Environmental Geology, 53, 915-922. Castellanza R., Lollino P., Ciantia M., 2018, A methodological approach to assess the hazard of underground cavities subjected to environmental weathering. Tunnelling and Underground Space Technology, 82, 278-292. Day A., 2002, Cave Surveying. British Cave Research Association, Cave Studies Series, 11. Del Prete S., Parise M., 2013, An overview of the geological and morphological constraints in the excavation of artificial cavities. In: Filippi M., Bosak P. (Eds.), Proc. 16th Int. Congress of Speleology, Brno, 21-28 July 2013, 2, 236-241. Delle Rose M., Federico A., Parise M., 2004, Sinkhole genesis and evolution in Apulia, and their interrelations with the anthropogenic environment. Natural Hazards and Earth System Science, 4, 747–755. Fazio N. L., Perrotti M., Lollino P., Parise M., Vattano M., Madonia G., Di Maggio C., 2017, A three-dimensional back analysis of the collapse of an underground cavity in soft rocks. Engineering Geology, 238, 301-311. Ferrero A. M., Segalini A., Giani G. P., 2010, Stability analysis of historic underground quarries. Computers and Geotechnics, 37 (4), 476-486. Fiore A., Parise M., 2013, Cronologia degli eventi di sprofondamento in Puglia, con particolare riferimento alle interazioni con l’ambiente antropizzato. Memorie Descrittive della Carta Geologica d’Italia, 93, 239-252. Fiore A., Fazio N. L., Lollino P., Luisi M., Miccoli N. M., Pagliarulo R., Perrotti M., Pisano L., Spalluto L., Vennari C., Vessia G., Parise M., 2018, Evaluating the susceptibility to anthropogenic sinkholes in Apulian calcarenites, southern Italy. In: Parise M., Gabrovsek F., Kaufmann G., Ravbar N. (Eds.), Advances in Karst Research: Theory, Fieldwork and Applications. Geol. Soc., London, sp. publ. 466, pp. 381-396. Fonseca C., 1980, Rupestrian civilization in Apulia. In: AA.VV., Apulia between Bisantium and the West. 36-116. Fraldi M., Guarracino F., 2009, Limit analysis of collapse mechanisms in cavities and tunnels according to the Hoek–Brown failure criterion. International Journal of Rock Mechanics and Mining Sciences, 46 (4), 665-673. Galeazzi C., 2013, The typological tree of artificial cavities: a contribution by the Commission of the Italian Speleological Society. Opera Ipogea, 1, 9-18. Guarino P. M., Nisio S., 2012, Anthropogenic sinkholes in the territory of the city of Naples (Southern Italy). Phys. Chem. Earth Parts A/B/C, 49, 92-102. Guarino P. M., Santo A., Forte G., De Falco M., Niceforo D. M. A., 2018, Analysis of a database for anthro- pogenic sinkhole triggering and zonation in the Naples hinterland (Southern Italy). Nat. Hazards, 91, 173–192. Gutiérrez F., Parise M., De Waele J., Jourde H., 2014, A review on natural and human-induced geohazards and impacts in karst. Earth Science Reviews, 138, 61-88. Häuselmann P., 2006, Symbols for karst surfaces: the UIS list. Zeitschrift fur Geomorphologie 147, International Atlas of Karst Phenomena, Sheets 18-21, 3-7. Häuselmann P. (Ed.), 2010, UIS Mapping Grades. Version 2. http://www.uisic.uis-speleo.org/UISmappingGrades.pdf Hutchinson D. J., Phillips C., Cascante G., 2002, Risk Considerations for Crown Pillar Stability Assessment for Mine Closure Planning. Geotechnical and Geological Engineering, 20 (1), 41-64. Klimchouk A. and Vjacheslav A., 2002, Karst breakdown mechanisms from observations in the gypsum caves of the Western Ukraine: implications for subsidence hazard assessment. International Journal of Speleology, 31, pp.55-88. Laureano P., 1993, Giardini di pietra. Bollati Boringhieri, Torino. Lollino P., Parise M., 2023, Sinkhole hazard quantitative assessment: Insights from the application of numerical modelling techniques. In: Land L., Kromhout C., Suter S. (Eds.), Proc. 17th Multidisc. Conf. on Sinkholes and Engng. and Environmental Impacts of Karst, Tampa (Florida, USA), 27-31 March 2023, NCKRI Symposium no. 9, pp. 141-150. Lollino P., Martimucci V., Parise M., 2013, Geological survey and numerical modeling of the potential failure mechanisms of underground caves. Geosystem Engng., 16 (1), 100-112. Palmer A.N., 2007, Cave Geology. Dayton, OH: Cave Books. Parise M., 2010, The impacts of quarrying in the Apulian Karst. In: Carrasco F., La Moreaux J.W., Duran Valsero J.J., Andreo B. (Eds.), Advances in Research in Karst Media. Berlin, Heidelberg, Springer, pp. 441-447. Parise, M., 2012, A present risk from past activities: sinkhole occurrence above underground quarries. Carbonates and Evaporites, 27 (2), 109-118. Parise M., 2015, A procedure for evaluating the susceptibility to natural and anthropogenic sinkholes. Georisk, 9, 272-285. Parise M., 2017, Engineering-geological studies in artificial cavities, aimed at evaluating the possibility of failures in underground settings. Proc. Int. Congress in Artificial Cavities “Hypogea 2017”, Cappadocia (Turkey), March 06-10, 2017, ISBN 978-605- 9680-37-0, pp. 137-144. Parise M., 2019, Sinkholes. In: White W.B., Culver D.C., Pipan T. (Eds.), Encyclopedia of Caves. Academic Press, Elsevier, 3rd edition, ISBN 978-0-12-814124-3, pp. 934-942. Parise M., 2022, Sinkholes, Subsidence and Related Mass Movements. In: Shroder J.J.F. (Ed.), Treatise on Geomorphology, vol. 5. Elsevier, Academic Press, pp. 200–220. https://dx.doi.org/10.1016/B978-0-12-818234- 5.00029-8. ISBN: 9780128182345.

64 A chronology of sinkholes related to artificial cavities in the hydrographic district of the Southern Apennines of Italy Parise M., Gunn J. (Eds.), 2007, Natural and Anthropogenic Hazards in Karst Areas: Recognition, Analysis and Mitigation. London Geol. Society, Special Publication 279. Parise M., Lollino P., 2011, A preliminary analysis of failure mechanisms in karst and man-made underground caves in Southern Italy. Geomorphology, 134 (1-2), 132-143. Parise M., Vennari C., 2013, A chronological catalogue of sinkholes in Italy: the first step toward a real evaluation of the sinkhole hazard. In: Land L., Doctor L.H., Stephenson B. (Eds.), Proc. 13th Multidisc. Conf. on Sinkholes and the. Engineering and Environmental Impacts of Karst, Carlsbad, National Cave and Karst Research Institute, pp. 383-392. Parise M., Vennari C., 2017, Distribution and features of natural and anthropogenic sinkholes in Apulia. In: Renard P., Bertrand C. (Eds.), EuroKarst 2016, Neuchatel. Advances in the hydrogeology of karst and carbonate reservoirs. Springer, ISBN 978- 3-319-45464-1, pp. 27-34. Parise M., Galeazzi C., Bixio R., Dixon M., 2013, Classification of artificial cavities: a first contribution by the UIS Commission. In: Filippi M., Bosak P. (Eds.), Proc. 16th Int. Congr. Speleology, Brno, 21-28 July 2013, 2, pp. 230-235. Pepe P., Pentimone N., Garziano G., Martimucci V., Parise M., 2013, Lessons learned from occurrence of sinkholes related to man-made cavities in a town of Southern Italy. In: Land L., Doctor L.H., Stephenson B. (Eds.), Proc. 13th Multidisc. Conf. Sinkholes Engng. Environ. Impacts of Karst, Carlsbad, National Cave and Karst Research Institute, pp. 393-401. Perrotti M., Lollino P., Fazio N. L., Pisano L., Vessia G., Parise M., Fiore A., Luisi M., 2018, Finite Element–based stability charts for underground cavities in soft calcarenites. ASCE International Journal of Geomechanics, 18 (7), 04018071. Swedzicki T., 2001, Geotechnical precursors to large-scale ground collapse in mines. International Journal of Rock Mechanics and Mining Sciences, 38 (7), 957-965. Tufano R., Guerriero L., Corona M. A., Bausilio G., Di Martire D., Nisio S., Calcaterra D., 2022, Anthropogenic sinkholes of the city of Naples, Italy: an update. Nat. Hazards, 1-32. Varriale R., Parise M., Genovese L., Leo M., Valese S., 2022, Underground Built Heritage in Naples: From Knowledge to Monitoring and Enhancement. In: D’Amico S., Venuti V. (Eds.), Handbook of Cultural Heritage analysis. Springer, pp. 2001-2035. Vennari C., Parise M., 2022, A chronological database about natural and anthropogenic sinkholes in Italy. Geosciences, 12, 200, https://doi.org/10.3390/geosciences12050200. Vennari C., Salvati P., Bianchi C., Casarano D., Parise M., Basso A., Marchesini I., 2022, AReGeoDatHa: Apulian regional GeoDatabase for geo-hydrological hazards. Journal of Environmental Management, 322, 116051. Waltham T., Bell F., Culshaw M., 2005, Sinkholes and Subsidence. Chichester, Springer.

65 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1 Dipartimento di Ingegneria, Università degli Studi della Campania “Luigi Vanvitelli”, Aversa, Italy 2   Dipartimento di Scienze della Terra e Geoambientali, Università degli Studi di Bari Aldo Moro, Bari, Italy 3   Institute of Research for Geo-Hydrological Protection, National Research Council (CNR-IRPI), Bari, Italy 4   Autorità di Bacino Distrettuale dell’Appennino Meridionale, Caserta, Italy *  Reference author email: [email protected] The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy: findings, architectural variability and risk assessment Maria Assunta Fabozzi1 , Isabella Serena Liso2 , Mario Parise2 , Carmela Vennari3 , Piernicola Lollino2 , Marco Vigliotti1 , Gennaro Capasso4 , Vera Corbelli4 , Daniela Ruberti1,* Abstract The present paper presents the preliminary results of a study project launched on the territory of one of the largest hydrographic districts in Italy, the Hydrographic District of Southern Apennine (Autorità di Bacino Distrettuale dell’Appennino Meridionale), in southern Italy, in collaboration with the Earth and Environmental Sciences Department at the University of Bari Aldo Moro and the Engineering Department of the University of Campania L. Vanvitelli. The objective of the research is the acquisition of all available information on the presence of cavities of anthropic origin in the subsoil of the District territory, in order to provide guidelines for the analysis and characterization of artificial cavities. An accurate research was conducted of all the documentary sources that provided information on the presence of cavities. The great heterogeneity of the data acquired made it necessary to identify a homogenization and cataloging system to be connected to the classification provided by the National Commission on Artificial Cavities of the Italian Speleological Society, and which takes into account the construction typologies and those structural elements susceptible to trigger phenomena of sinkholes. Since the municipalities included in the District are 1632, a “top-down” approach was chosen for the identification of the presence of possible anthropic cavities, starting from the analysis of the Lithological Map of Italy and selecting those lithologies potentially affected by mining activities. This has reduced the number of administrative areas to check. Although the work is still preliminary, it will constitute a single document in Italy and will allow for the definition of procedures for preliminary evaluation of the stability conditions of artificial cavities. Keywords: artificial cavities, inventory, typology, risk. Introduction Sinkholes phenomena induced by the widespread presence of anthropogenic cavities are frequent and well-known in Italy (Parise, 2019, 2022). Nevertheless, in many urban centers, cavities have been reported in specific geological investigations, but the real extent of underground voids is often known only partially, and this might have serious effects on the built-up environment. In these towns, artificial cavities have been built for different purposes such as hydraulic, religious, military, transit works (Palmer, 2007; Parise, 2007, 2009, 2012b; Del Prete and Parise, 2013); among the others, underground mining activities were very common, and excavated to extract rock material for buildings. Later on, the urban development has sealed every signal of the presence of cavities, bringing the local communities to loss of memory about them, and nowadays they represent a geological hazard that significantly contributes to subsoil instability at many places (Del Prete et al., 2010; Parise and Lollino, 2011; Parise 2012a, 2017). The need to survey underground artificial cavities in urban centers derives from two reasons. First, the anthropic hypogea represent an absolute documentary value, still unduly neglected and little used for the purposes of a correct and sustainable management of the territory, including its natural resources and historical and artistic heritage. The enhancement and sustainable reuse of hypogea could contribute to enhance the cultural and tourist promotion of many territories, especially in small municipalities. Furthermore, in a correct urban management, knowledge of the subsoil is a priority, as the presence of cavities may easily be at the origin of underground failures, up to sinkhole development. With the aim of acquiring as much information on the presence of underground cavities and provide

66 The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy guidelines for the analysis and characterization of artificial cavities, a study project was launched on the territory of one of the largest hydrographic districts in Italy, the Hydrographic District of Southern Apennine (Autorità di Bacino Distrettuale dell’Appennino Meridionale, ADAM from now on), in southern Italy, in collaboration with the Earth and Environmental Sciences Department at the University of Bari Aldo Moro and the Engineering Department of the University of Campania L. Vanvitelli. The specific goals of this research study can be summarized as follow: i) the recognition of the largest number of cavities throughout the territory; ii) the highlight of the main extraction characteristics (architecture) for the cavities related to extraction works; iii) eventually, the definition of guidelines for their characterization, aimed at providing support for the management of the hazard related to presence of artificial cavities. Study area The ADAM territory (fig. 1) – as defined by art. 64 of Legislative Decree no. 152/2006 (implementing Directive 2000/60/CE) – includes the whole Regions of Basilicata, Calabria, Campania, Apulia and part of Abruzzo, Latium and Molise, comprising 25 Provinces (6 partially), 1632 Municipalities, 100 Mountain Communities, 44 Reclamation Consortia, 978 Protected Natural Areas, with a resident population of almost 14 million inhabitants, which represents about 23% of the national population. The ADAM territory is characterized by a significant development of coastal areas (about 2,100 km) and is touched by the Tyrrhenian Sea to the west, the Adriatic Sea to the east, and the Ionian Sea to the southeast and the south. The landscape has a heterogeneous morphology, from mountainous to hilly, with large alluvial and intermontane plains. The Apennine chain, crossing the District from north to south, acts as a TyrrhenianAdriatic watershed and comprises the Southern Apennines and part of the Abruzzese ones. The Apennine Chain is mainly composed of limestone dissected by plateaus and basins parallel to the ranges (Calamita et al., 2009; Beltrando et al., 2010; Vezzani et al., 2010; Bucci et al., 2021). The ridges are formed by Mesozoic carbonate rocks and marly arenaceous-clayey soils; only the Sila and Aspromonte chains are made up of metamorphic and igneous granite rocks. The slopes of the Apennine valleys are often affected by landslides due to the predominant clayey-marly nature of the soils (Loche et al., 2022). Furthermore, in northern Campania, the presence of three volcanic districts (Roccamonfina, Somma-Vesuvius and Phlegrean Fields) has produced a large distribution of lava and volcaniclastic products, both at the surface and in the subsoil (De Vivo et al., 2019). A volcanic district is known also in northern Basilicata (Mt. Vulture; Schiattarella et al., 2005). Fig. 1 – Location map of the ADAM territory.

67 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa The complex geological history of the Apennine chain and, therefore, of the geometric relationships between the various stratigraphic-structural units results, in the study area, into a considerable variability of the lithological and permeability characteristics, conditioning the distribution and geometry of the hydrogeological structures and the pattern of groundwater circulation at both small and large scale. Identification of the hypogea The first step in mitigating the sinkhole risk is obviously the accurate knowledge of the cavities, and therefore their punctual inventory, with GIS mapping and digitized cartography. With the aim of recovering as much information as possible on the presence of artificial cavities, a study was carried out aimed at updating the list of municipalities affected by the presence of cavities of anthropogenic origin, falling in the area under ADAM’s jurisdiction. The first approach was to analyze the ADAM database (DB) to recover the Urban Municipal Plans (Piani Urbanistici Comunali), where implemented, and to verify the possible indication of cavities in the subsoil of the urban centers (fig. 2). When these documents are not available (either because they were never produced, or because they are not available at ADAM), reference was made to the DBs of the two university research centers collaborating in the project, and deriving from specific study activities conducted by the researchers involved. Another reference DB is the classification provided by the National Commission on Artificial Cavities of the Italian Speleological Society (Società Speleologica Italiana, SSI; Galeazzi, 2013), later adopted by the International Union of Speleology (UIS; Parise et al., 2013), based on data that define topography, morphometry and category of each cavity, as it will be discussed later. From the joint analysis it was thus possible, in a first phase, to identify and select the areas most affected by the presence of cavities. In both cases, it is worth to mention the extreme variability of the available data (points, areas, maps), which requires a great effort in putting them together in the attempt of homogeneization. However, such an approach to consult urban plans requires a long time, being 1632 the municipalities falling within the ADAM territory. It was therefore considered appropriate to carry out a parallel “top-down” approach to select the municipalities potentially affected by the presence of cavities: starting from the Geological Map of Italy (fig. 3; Bonomo et al., 2005), the lithologies potentially affected by mining activities were initially selected. In GIS environment, a query on the attributes was created, which made it possible to select the areas characterized by the selected lithologies; subsequently, a topological overlay with the administrative limits returned the municipal areas potentially affected by cavities (fig. 2, 3). This made possible to enormously reduce the number of documents to examine. Finally, further indications about the probable presence of underground cavities can be deduced by consulting the catalogue of anthropogenic sinkholes populated since several years by a joint project between the Institute of Research for Geo-Hydrological Protection of the National Research Council of Italy and the Earth and Environmental Sciences Department at Aldo Moro University in Bari (Parise and Vennari, 2013; Vennari and Parise, 2022; see Liso et al., this volume). By extracting from this catalogue the events of sinkholes related to artificial cavities that occurred within the Hydrographic District of Southern ApenFig. 2 – Flow chart of the research activities.

68 The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy Fig. 3 – Geolithological map of the ADAM territory (modified from Bonomo et al., 2005). Key: 1) Basalts and ofiolites; 2) Cristalline igneous and metamorphic rocks; 3) Volcaniclastic deposits; 4) Lavas and piroclastic products; 5) Clayey-arenaceous-marly deposits; 6) Marly limestone with chert; 7) Detrital bioclastic limestone (“panchina”); 8) Micritic and bioclastic limestone; 9) Dolomite; 10) Alluvial terraced clay-silt and sands; 11) Lacustrine clay and silt; 12) Aeolian sands; 13) Sandstones and conglomerates; 14) Clay; 15) Gessoso-Solfifera Formation; 16) travertine; 17) lakes.

69 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa nines, it is possible to indirectly obtain indications on the presence of artificial cavities where there was no previous indication. Architecture and classification of the cavities Since 1928, the awareness of the large diffusion of natural hypogea throughout the Italian territory has pushed the creation of a classification, provided by the National Commission on Artificial Cavities of the SSI (Galeazzi, 2013), later adopted by the UIS (Parise et al., 2013). The classification defines and codifies the typology according to the function for which an artificial cavity has been realized (fig. 4). Given the wide variety of structures, the classification is organized as a tree, based on seven main types, identified with a capital letter, in turn divided into sub-types identified by a number according to the main purpose of original construction of the cavity (Parise et al., 2013; fig. 4). It should also be noted that in recent times many excavations have been conducted underground for infrastructures connected to modern civilization (including subways, car parks, tunnels, military works, and so on). However, the artificial cavities have been made for thousands of centuries; therefore, alongside the classification proposed in figure 4, a catalogue has been created which indicates the age of the first construction of the cavity (from prehistoric to the 20th century, and later). The age is accompanied by information on the techniques and purpose of the excavation, shapes and sizes, and any other correlation with historical events. Of significant importance are above all the construction techniques (Varriale et al., 2022; Bixio et al., 2023): the cavities could be dug or constructed in the subsoil or they could be natural caves modified by men; in other cases, they resulted from the burial of structures originally located on the surface (Parise et al., 2013). In large portions of the ADAM territory the presence of artificial cavities is reported, differing in excavation techniques, shape and size according to the use and, above all, to the extracted lithotypes. On the northern Tyrrhenian side, hypogea are mainly obtained from the volcanic and volcaniclastic deposits of the main volcanic districts of northern Campania, i.e. Somma-Vesuvius, Roccamonfina and Phlegrean Fields. Above all, the latter are responsible for the settling of massive deposits of tuff and loose pyroclastics widely used, for thousands of years, as building materials (Vigliotti and Ruberti, 2018). Mining volcanic tuff was conditioned by the morphoFig. 4 – Typological tree for the classification of artificial cavities (from Parise et al., 2013).

70 The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy Fig. 5 – Some types of artificial cavities in northwest Campania Region; B4 – warehouses: a) at San Marcellino (restored), b) at Sant’Arpino (abandoned), c) at Casal di Principe (restored), d) at Sant’Arpino (restored) with F4 – shaft; e) at Aversa (restored; now B9 - other: underground parking); f) Santa Maria in grotta, C1 – place of worship at Sessa Aurunca (photo courtesy Alessandra Ruberti).

71 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa logical setting: on slopes, where the tuff was outcropping, excavation proceeded in the rock wall; in lowland areas, where the tuff lies under loose volcanic and alluvial materials, it was initially reached from above through vertical wells (drills), until reaching and exceeding the roof of the tuffaceous bank; then the cultivation went on through chambers and/ or single or multiple tunnels of various development which could later be reached through stairs or ramps, depending on the depth (fig. 3). Even the materials above the tuff (pozzolana, scoriae, lapilli) were extracted on site, underground, by creating large cavities which were covered as the extraction progressed, or through narrow tunnels arranged in a radial pattern, starting from a central well which formed the front of attack. The presence of voids is not only linked to mining activities: a dense network of tunnels constituted a water network, in which the water that flowed over the surface fed underground rainwater cisterns. Where these tunnels were supplying with water the main inhabited areas, they became underground aqueducts (Bixio et al., 2000, 2007), and some of them are several kilometers long, with impressive structures, still functioning nowadays. Some hypogea were used as places of worship, others as cellars for the conservation of perishable foods and, more recently, before being completely abandoned, as anti-aircraft shelters during the Second World War (fig. 5). Subsequently, many cavities were partially filled by dumping waste materials and solid and liquid waste from the wells, obliterating their presence and becoming a source of danger. In peculiar geomorphological situations, such as the flanks of incised valleys (called gravine in Apulia; see Parise et al., 2003), civilian settlements were established, following the rupestrian civilization that characterized wide sectors of southern Italy during the Middle Age: these cavities produced a complex network of voids, often linked one to the others, and distributed over multiple levels (fig. 6). The wide diffusion of artificial cavities characterized by high architectural variability throughout the ADAM territory, and the large quantity of extremely heterogeneous data acquired, required the creation of a geodatabase that took into account the peculiarities of the local/regional sites, integrating the catalogue produced by SSI which defines the main historical characteristics of each hypogeum. A first step (requirement analysis) involved the identification of the information to be managed and the relationships between them, which was followed by a phase during which the relative entities and relationships were defined (conceptual planning); during the logical design, the tables and the relationships were defined, and finally the database was physically built. The subsequent phases of Population (data entry), Testing and Management (applying filters and building queries) made it possible to verify its functionality. Starting from the National Register of Artificial Cavities, produced by SSI, which defines a basic level of knowledge (plano-altimetric location, volume, lithology, past and current uses, state of conservation), the database created (fig. 7) provides for the collection of data of each part (chambers, wells, descents) of a cavity system, and the indication of the presence of any instability and/or highlighting those elements that can cause instability phenomena. This DB can be implemented by a photographic report, plans, sections and 3D models for each hypogeum, when available. Conclusive remarks Within the framework of a project funded by the Hydrographic District of the Southern Apennines, we have presented in this work a preliminary survey of the cavities in the ADAM area, and the survey strategies adopted. The extreme variability of the artificial cavities was highlighted, in terms of age, type, use and state of conservation. The work is still in progress and many portions of the territory are still without data, or need to verify what has been identified with the survey framework set out herein. However, the acquired data is populating a DB which will constitute a unique document in Italy, including all aspects relating to artificial cavities which will form the basis for subsequent analyses, aimed at defining procedures for preliminary evaluation of the stability conditions of artificial cavities, detailed assessment at two pilot areas within the ADAM territory, including modeling analyses, and to establishment of actions addressed toward mitigation of the risk related to sinkholes. Acknowledgments This work was partly carried out within the framework of two joint contracts between the Hydrographic District of Southern Apennines and, respectively, the Earth and Environmental Sciences Department at University Aldo Moro in Bari (scientific responsible: M. Parise) and the Engineering Department at University “Luigi Vanvitelli” in Aversa (scientific responsible: D. Ruberti).

72 The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy Fig. 6 – Some types of artificial cavities in Apulia Region: a) A.1 – Underground water ducts: aqueducts at Supersano (photo courtesy Francesco De Salve); b) Fonte Pliniano, A.5 – wells collecting freshwater springs at Manduria; c) B.1 – Permanent dwellings at Massafra; d) B3 – Oil mill factory at Alessano; e) Grotta di San Michele, C1 – place of worship at Ceglie Messapica; f) E.1 – Aggregate quarries at Cutrofiano.

73 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Bibliography Beltrando M., Peccerillo A., Mattei M., Conticelli S. and Doglioni C., 2010, The Geology of Italy. Journal of the Virtual Explorer, 36, paper 33. doi:10.3809/jvirtex.2010.00256, 2010 Bixio R., Castellani V., Maifredi P., Saj S., 2000, L’acquedotto sotterraneo di Gravina in Puglia “Sant’Angelo-Fontane della Stella”, in Il Parco della Pietra e dell’Acqua: pp. 215-253. Consorzio Sidin/UNESCO, Gravina in Puglia. Bixio R., De Pascale A., Galeazzi C. & Parise M., 2023, Rupestrian works and artificial cavities: categories of construction techniques. Journal of Architectural and Engineering Research 4: pp. 59-68. Bixio R., Parise M., Saj S., Traverso M., 2007, L’acquedotto sotterraneo di Gravina in Puglia “Sant’Angelo-Fontane della Stella”. Opera Ipogea, anno IX (1): pp. 105-112. Fig. 7 – Structure of the main DB table with indication of the fields and the type of value allowed for a database attribute.

74 The hidden world of artificial cavities in the hydrographic district of the Southern Apennines of Italy Bonomo R., Capotorti F., D’Ambrogi C., Di Stefano R., Graziano R., Martarelli L., Pampaloni M.L., Pantaloni M., Ricci V., Compagnoni B., Galluzzo F., Tacchia D., Masella G., Pannuti V., Ventura R., Vitale V., 2005, Carta geologica d’Italia alla scala 1:1.250.000, Serv. Geol. d’It., APAT, Roma. Bucci F., Santangelo M., Fongo L., Alvioli M., Cardinali M., Melelli L., Marchesini I., 2021, A new digital lithological Map of Italy at 1:100.000 scale. PANGAEA, https://doi.pangaea.de/10.1594/PANGAEA.935673 (dataset in review) Calamita F., Esestime P., Paltrinieri W., Scisciani V., Tavarnelli E., 2009, Structural inheritance of pre- and synorogenic normal faults on the arcuate geometry of Pliocene-Quaternary thrusts: Examples from the Central and Southern Apennine Chain. Italian Journal of Geosciences (Boll. Soc. Geol. It.), 128, 2, 381-394 (DOI:10.3301/IJG.2009.128.2.381) De Vivo B., Belkin H., Rolandi G. (Eds.), 2019, Vesuvius, Campi Flegrei, and Campanian Volcanism. Elsevier, 520pp., ISBN:9780128164549 Del Prete S. & Parise M., 2007, L’influenza dei fattori geologici e geomorfologici sulla realizzazione di cavità artificiali. Opera Ipogea, anno 9, n. 2, p. 3-16. Del Prete S., Galeazzi C., Germani C., Martimucci V., Parise, M., 2010, Geohydrological Problems Related to Underground Artificial Cavities. In Proceedings of the Proceedings of the Convegni Lincei “Frane e dissesto idrogeologico: Consuntivo”; Accademia dei Lincei, Rome, Italy, 22/03/2010; pp. 223–232. Galeazzi C., 2013, The typological tree of artificial cavities: a contribution by the Commission of the Italian Speleological Society. Opera Ipogea, 1, 9–18. Liso I.S., Vennari C., Fabozzi M.A., Ruberti D., Vigliotti M., Capasso G., Corbelli V., Parise M., this volume, A chronology of sinkholes related to artificial cavities in the Hydrographic District of the Southern Apennines of Italy. Loche M., Alvioli M., Marchesini I., Bakka H. and Lombardo, L., 2022, Landslide susceptibility maps of Italy: Lesson learnt from dealing with multiple landslide types and the uneven spatial distribution of the national inventory. Earth-science reviews, 1-21. [104125].https://doi.org/10.1016/j.earscirev.2022.104125 Palmer A., 2007, Cave geology. Cave Books. Parise M., 2007, Il Progetto “La Carta degli Antichi Acquedotti Italiani”. Opera Ipogea, anno 9, no. 1, 3-16. Parise M., 2009, Distribution and characteristics of ancient underground aqueducts in Italy. IWA Specialty Conference, 2nd Int. Symp. “Water and wastewater technologies in ancient civilizations”, Bari, 28-30 May 2009. Parise M., 2012a, A present risk from past activities: sinkhole occurrence above underground quarries. Carbonates and Evaporites, 27 (2), 109–118 Parise M., 2012b, Lo sviluppo degli insediamenti rupestri in funzione delle caratteristiche fisiografiche del territorio. Atti Seminario Internazionale CRHIMA “Rupestrian settlements in the Mediterranean region”, Massafra, Aprile-Maggio 2011, ISBN 978-88- 96080-06-1, p. 133-138. Parise M., 2017, Engineering-geological studies in artificial cavities, aimed at evaluating the possibility of failures in underground settings. Proc. Int. Congress in Artificial Cavities “Hypogea 2017”, Cappadocia (Turkey), March 06-10, 2017, ISBN 978-605- 9680-37-0, pp. 137-144 Parise M., 2019, Sinkholes. In: White W.B., Culver D.C. & Pipan T. (Eds.), Encyclopedia of Caves. Academic Press, Elsevier, 3rd edition, ISBN 978-0-12-814124-3, 934-942. Parise M., 2022, Sinkholes, Subsidence and Related Mass Movements. In: Shroder J.J.F. (Ed.), Treatise on Geomorphology, vol. 5. Elsevier, Academic Press, pp. 200–220. https://dx.doi.org/10.1016/B978-0-12-818234- 5.00029-8. ISBN: 9780128182345. Parise M., Federico A., Delle Rose M., Sammarco M., 2003, Karst terminology in Apulia (southern Italy). Acta Carsologica 32 (2): pp. 65-82. Parise M., Galeazzi C., Bixio R., Dixon R., 2013, Classification of artificial cavities: a first contribution by the UIS Commission. In: Filippi, M. and Bosak, P. (Eds.), Proc. 16th Int. Congr. Speleol., 21-28 July 2013, 2, 230-235 Parise M., Lollino P., 2011, A Preliminary Analysis of Failure Mechanisms in Karst and Man-Made Underground Caves in Southern Italy. Geomorphology, 134, 132–143, doi:10.1016/j.geomorph.2011.06.008. Schiattarella M., Beneduce P., Di Leo P., Giano S.I., Giannandrea P., Principe C.,2005, Assetto strutturale ed evoluzione morfotettonica quaternaria del vulcano del Monte Vulture (Appennino lucano). Boll. Soc. Geol. It., 124: 543-562. Varriale R., Parise M., Genovese L., Leo M. & Valese S., 2022, Underground Built Heritage in Naples: From Knowledge to Monitoring and Enhancement, in D’Amico S., Venuti V. (eds), Handbook of Cultural Heritage analysis, pp. 2001-2035. Springer. Vezzani L., Festa A., Ghisetti F.C., 2010, Geology and tectonic evolution of the central-southern Apennines, Italy, Geological Society of America Special Paper, 58pp., doi: 10.1130/2010.2469 Vigliotti M.; Ruberti D., 2018, Campania Grey Tuff and Anthropogenic Tuff Cavites in the Southern Metropolitan Area of Caserta (Southern Italy). Alp. Mediterr. Quat., 31, 247–251.

75 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Department of Earth and Environmental Sciences, University of Bari Aldo Moro, Italy *  Reference Author: [email protected] On the role of geo-structural features in the development of failure mechanisms affecting manmade underground cavities Piernicola Lollino1,*, Mario Parise1 Abstract The stability of man-made underground caves is progressively challenging throughout Italy due to the increase in urbanization processes as well as the gradual environmentally-driven degradation processes affecting caves excavated several decades ago and, later on, abandoned. In general, such processes interest the caves as a consequence of prolonged high-humidity conditions or, under more critical conditions, sudden inflow of a large amount of water due to pipeline leakages or rainfall infiltration. In the last years, a strong advancement in the 2-D and 3-D numerical modelling aimed at investigating the stability of underground caves within soft rock masses has been developed, and several scientific articles have been published in this perspective. In most of these works, the rock mass is simulated as continuous, consistently with the absence of relevant fractures or discontinuity sets, as largely observed within these geological settings. However, in some cases persistent discontinuities can affect the rock masses hosting the caves, and, accordingly, such geo-structural conditions need to be properly accounted for in assessing the safety of the underground environments. In this contribution, the role of joint inclination in the development of failure mechanisms within anthropogenic caves is firstly investigated by means of two-dimensional parametric finite element analyses of an ideal underground cave implementing discrete joints with different inclination. From a numerical point of view, the influence of discontinuities potentially existing in the calcarenite masses is explored by implementing joints as interfaces between continuous rock domains. In particular, numerical models with vertical and inclined joints are assumed and the corresponding results highlight the impact of the joint dip in the generation and enhancement of the failure mechanisms within the rock masses, as a function of the joint shear strength. Keywords: artificial cavities, stability, failure, weathering, modeling. Introduction The potential failure of man-made underground caves, that were exploited and abandoned decades ago, is still nowadays frequently underestimated in land and urban planning policies, and has often serious consequences on the built-up environment. For artificial cavities within soft, porous rock masses, water infiltration from the ground surface or from pipeline leakages, water inflow into the caves, change in the ventilation conditions, and other factors, can be responsible for the degradation of the rock material properties over time, even in a relatively short time. Recent case studies regarding the failure of man-made underground cavities, which caused sinkholes affecting urbanized areas, are reported in previous works, as for example the sinkholes affecting calcarenite quarries in Southern Italy (Parise 2010, 2015; Vattano et al., 2013), mining caves in Canada (Bétournay, 2009), the siltstone Longyou caverns in China (Li et al., 2009), and the limestone mines in the Netherlands and Belgium (Van Den Eeckhaut et al., 2007). In the last years, numerical modelling has represented a highly efficient tool to investigate the stress-strain evolution of the rock mass around the cave, so that the variations of equilibrium and the associated displacement field generated by a change in loading condition, or in boundary conditions, can be accurately simulated (Ferrero et al., 2010; Parise & Lollino, 2011; Zhang et al., 2016; Fazio et al., 2017; Castellanza et al., 2018; to mention a few). However, such numerical applications, which are basically performed with the finite element method (FEM), have mainly focused on intact rock conditions. As such, the rock mass is simulated as continuous, consistently with the absence of relevant fractures or discontinuity sets, as largely observed within these geological settings. However, in some cases persistent discontinuities can affect the rock masses hosting the caves, and, accordingly, such geo-structural conditions need to be properly accounted for in the safety assessment of the underground environments (Parise, 2012). Nowadays, the presence of joints within FEM analyses can be taken into account in a relatively easy way by employing interface or joint elements devised to represent discontinuities (Hammah et al., 2008; De Silva et al., 2022), although the main shortcoming existing in this application is represented by aperture and sliding possibility within the rock mass not being allowed, due to the continuity assumption made in the same methodology. In this

76 On the role of geo-structural features in the development of failure mechanisms affecting man-made underground cavities perspective, an important opportunity is provided by discrete element methods (DEM; Lollino et al., 2004) or the more recent hybrid finite element/discrete element methods (FEM-DEM; Munjiza, 2004; Lollino & Andriani, 2017; Lollino & Parise, 2023). The present paper is aimed at proposing a preliminary investigation on the role of structural features, as discrete joints or proper joint sets, in the equilibrium state of artificial cavities by means of finite element analyses carried out on an ideal underground cave implementing discontinuities within the rock mass domain. To this purpose, an overview of the geological background and typical structural features of the region under study is firstly outlined. Then, a parametric analysis made up of different two-dimensional numerical models characterized by different joint inclination angle values is proposed, and some interesting insights in the failure mechanism of underground caves are highlighted. Geological background and typical structural features Apulia, the south-eastern heel of the Italian boot, widely presents outcrops of calcarenite rocks of different ages, from Miocene to upper Pleistocene, that have been affected by excavation, given the soft, or very soft, nature of the rock, making them easy to work, but at the same time guaranteeing sufficient strength to support excavations (Fiore et al., 2018). The different calcarenites may be highly variable in terms of their petrographic and structural features, and this has effects on the mechanical response of the rock, leading to a possible high degree of heterogeneity at different sites. In particular, the increase in water, eventually up to approaching the saturation point, is among the main factors potentially bringing the calcarenites close to general failure. Overall, the physical properties of the rock masses allowed to develop many different typologies of artificial cavities, from civilian settlements to underground aqueducts and cisterns, to olive mills up to quarries (figure 1). In most of the cases, calcarenites can be ascribed to a continuous rock mass, due to the very scarce presence of fractures and joints, that are typically concentrated close to the cavity entrances, or nearby the open slopes, as an effect of stress release. However, in some conditions the calcarenite rock mass may locally present joints that appear to control the likely failures, both as isolated joints and sets or families. In such a situation, the role played by these discontinuities cannot be neglected and should therefore be included in the modelling analysis. Numerical analyses In order to investigate the role of structural features in the equilibrium of the rock mass forming the cavity roof, a two-dimensional finite element model of an ideal underground cave characterized by a single joint set, with variable inclination angle and shear strength properties, has been developed. The model sizes are 50 m x 30 m, with an inner rectangular region subjected to excavation representing the cavity, 10 m wide and 6 m high, with roof thickness equal to 6 m. Conventional boundary conditions typically used in static analyses, with null horizontal and vertical displacement imposed along the model bottom and null horizontal displacement prescribed along the vertical boundaries, have been applied to the model. Free displacements have been imposed at the top boundary representing the ground surface. According to the model geometry described, the external boundaries result to be sufficiently far from the process zone, i.e. the area around the cavity, so that the risk of numerical boundary effects is significantly reduced. The adopted discretization mesh is represented in figure 2, which shows a coarser mesh in the domain areas far from the process zone, and a significantly finer mesh in the area around the cavity, where high strain levels are supposed to be expected. Therefore, the mesh proposed represents a reasonable compromise between significantly enhanced computational accuracy and relatively short computational time. A single joint set, with variable joint inclination, persistence and shear strength, has been simulated in the different analyses to explore the role of discontinuities in the equilibrium conditions of the rock mass. Rock properties representative of the typical mechanical behavior of the soft calcarenite outcropping in Apulia, where such anthropogenic cavities have been excavated, as described in previous studies performed by the authors (Perrotti et al., 2018), have been impleFig. 1 – Picture of a typical underground cave in soft calcarenite rocks.

77 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa mented for the rock matrix. To simulate the rock mass behaviour, an elastic-perfectly plastic model, with a Mohr-Coulomb failure criterion, has been assumed. The rock material properties are summarized in Table 1. For the rock joints, a shear strength, characterized by null cohesion and tensile strength, as well as variable friction angle values, has been chosen in the analyses. As such, the shear strength is represented uniquely by the friction angle, whereas the elastic normal and shear stiffness values of the joints have been assumed to be, respectively, equal to: Jkn = 100000 MPa/m, jks = 10000 MPa/m. An initial stress state resulting from geostatic vertical stresses and horizontal stresses derived from a k0 = 1 assumption, i.e. σh = σv , has been simulated in the analyses. In order to investigate the role of joint inclination with respect to the horizontal direction (α value in fig. 2), different values of the same parameter have been assumed in the models, ranging from vertical joints (α = 90°) to inclination angle of α = 65°, 45°, 30° and 20°. A joint spacing equal to s = 2 m has been kept as fixed in all analyses here described, whereas an infinite persistence has been also assumed. For the different models implementing variable joint inclination, different values of the joint shear strength have been assumed in order to detect the strength values mobilized at failure conditions. In particular, friction angle values ranging from ϕ’ = 32°, which corresponds to highly smooth surface joints (i.e. no roughness and no dilation), to ϕ’ = 55°, corresponding to significantly rough joints (i.e. large dilation), have been taken into account. Model instability has been defined when a clear failure mechanism, characterized by a well-defined concentration of plastic shear strains or plastic zones as well as by increased displacement in the area delimited by the plastic zones, is detected and, also, numerical convergence of the analysis is not reached. Fig. 2 – Numerical domain and discretization mesh adopted. Property γ (kN/m3 ) E (MPa) ν c’ (kPa) ϕ (°) σt (kPa) Value 17 200 0.3 200 30 200 Table 1 – Material properties adopted for the calcarenite rock.

78 On the role of geo-structural features in the development of failure mechanisms affecting man-made underground cavities Results and discussion A preliminary model, with no joints (i.e. intact rock mass) has been carried out and the results indicate stable conditions of the cavity roof. The numerical results corresponding to the model with vertical joints indicate a strong influence of the joint shear strength in the equilibrium conditions of the cavity roof. In particular, when joint friction angle values larger than ϕ’ = 38° are adopted, the model does not show any sign of instability. With ϕ’ = 36°, the model is in a marginal stability state, characterized by a straining mechanism that involves an area corresponding to the whole cavity roof, thus resembling the typical mechanism of collapse sinkhole (Gutierrez et al., 2014; Parise 2019, 2022). However, under these assumptions, the model is still stable from a numerical point of view, i.e. numerical convergence is still ensured. When a lower value of the joint friction angle is taken into account (ϕ’ < 35°), the model reaches numerical instability, with a typical chimney failure mechanism delimited by the joints aligned with the vertical boundaries of the cavity (fig. 3). In particular, figure 3a shows the contours of the calculated plastic shear strains, which indicate concentrated plastic zones at the upper and lower extremes of the joints, whereas the cumulated total displacements, showing significant downward displacement values concentrated within the overall cavity roof, are reported in figure 3b. Therefore, the above described model results to be highly sensitive to the joint shear strength parameter, since the equilibrium conditions of the cave roof change from stable to unstable, according to limited variations in the same parameter. The analyses with inclined joints revealed relevant outcomes in terms of cave failure mechanism as well as shear strength threshold values for cave stability. In particular, when a large joint inclination angle is assumed (α = 65°), the model results to be stable for joint friction angle larger than ϕ’ = 44°. For lower values of the same parameter, a general failure, as well as a lack of numerical convergence, are calculated in the analysis. Therefore, the threshold friction angle value for stability is calculated to be approximately equal to ϕ’ = 44°. In figures 4a and 4b, in order to give evidence of the failure mechanism, the yielded zones and the total displacement contours calculated in the analysis assuming ϕ’ = 42° are plotted, respectively. Both figures highlight a clear tendency to develop a straining mechanism involving the whole cavity roof (i.e. up to the ground surface), which is confined by the highly-inclined joints. When a joint inclination angle equal to α = 45° is assumed, the model results to be stable (i.e. numerical convergence is reached) for joint friction angle larger than ϕ’ = 47°. For lower values of the same parameter, the tendency to develop triangular-shaped local failures along the upper boundary of the cavity, which are delimited by the inclined joints and new plastic zones developing within the intact rock, is simulated in the analysis. In figures 5a and 5b, the plastic shear strains and the total displacement contours calculated in the analysis assuming ϕ’ = 45° are plotted, respectively, as representative of the typical failure mechanism occurring in this model. Both figures clearly show the local failure developed along the cavity roof, although some straining effects are also calculated even in the upper area, i.e. up to the ground surface. In the case of a joint inclination angle equal to α = 30°, the model results to be stable (i.e. numerical convergence is reached) for joint friction angle larger than ϕ’ = 50°. For lower values of the same parameter, a triangular-shaped local failure, which is formed by the inclined joints and new yielded zones that develop within the intact rock, tends to develop along the upper boundary of the cavity. In figures 6a and 6b, the plastic zones and the total displacement contours calculated in the analysis assuming ϕ’ = 49° are plotted, respectively. The previous failure mechanism is strongly enhanced in the analysis with α = 20° joint inclination angle (fig. 7), for which a threshold joint friction angle for stability equal to ϕ’ = 50° has been obtained. In this case, the local failure along the cavity roof is clearly identified, but a straining mechanism involving the whole roof thickness up to the ground surface is generated as a consequence of the loosening effect of the local failure. A plot summarizing the threshold joint friction angle values for stability against the joint inclination angle within the different numerical models has been reported in figure 8. It indicates a clear trend of reduction in the threshold value for stability of the friction angle, from values larger than 50° to 35°, with joint inclination angle values increasing from α = 20° to α = 90°. These results indicate that, for highly inclined joints, rock mass instability can be triggered even with low friction angle values of the joints, whereas for low-inclined joints cavity failure can be triggered only with very large friction angle values. These outcomes strongly highlight the need of a proper geological and structural characterization of the rock masses, that appears to greatly influence the possibility of reaching failure conditions. The numerical results above described also show that a clear change in the failure mechanism arises when the joint inclination changes, since with highly-inclined joints the tendency to proper sinkholes (chimney-shaped mechanisms) dominates (Berest, 2017), whereas with low-inclined joints local triangular-shaped failures take place along the upper boundary of the cavity. In this latter case, the possibility to propagate upward, until reaching the ground surface and producing a collapse sinkhole, seems to be lower. Models implementing non-persistent joint sets have been also developed and the corresponding results indicate a strong influence of the joint persistence in the equilibrium state of the rock mass. In particular, in the case of non-persistent joints, the presence of rock bridges significantly reduces the tendency to instability of the rock mass above the cavity.

79 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Fig. 3 – Numerical results with vertical joints: a) contours of plastic shear strains; b) cumulated displacements. Fig. 4 – Numerical results with α = 65° inclined joints: a) plastic zones; b) cumulated displacements.

80 On the role of geo-structural features in the development of failure mechanisms affecting man-made underground cavities Fig. 5 – Numerical results with α = 45° inclined joints: a) contours of plastic shear strains; b) cumulated displacements. Fig. 6 – Numerical results with α = 30° inclined joints: a) plastic zones; b) cumulated displacements.

81 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Concluding remarks This paper has investigated the role of joint sets, with large spacings, within the vaults of anthropogenic caves in soft calcarenite rock masses. To this purpose, numerical models with vertical and inclined joints have been assumed and the corresponding results highlight the impact of the joint dip in the generation and enhancement of the failure mechanisms within the rock masses, as a function of the joint shear strength. The analyses are still on going, taking advantage of the large amount of data regarding artificial cavities in calcarenites of Apulia, and will further proceed in delineating the other factors that, together with the physical properties and the geo-structural features, may contribute to the development of underground failure mechanisms. Fig. 7 – Numerical results with α = 20° inclined joints: failure mechanism. Fig. 8 – Plot of threshold joint friction angle values for stability against joint inclination angle.

82 On the role of geo-structural features in the development of failure mechanisms affecting man-made underground cavities Acknowledgments This work was partly carried out within the framework of the project RETURN (multi-Risk sciEnce for resilienT commUnities undeR a changiNg climate), project code MUR: PE00000005, in the Spoke 2 VS2-Ground instabilities. Bibliography Berest P., 2017, Cases, causes and classifications of craters above salt caverns. International Journal of Rock Mechanics and Mining Sciences 100, pp. 318-329. Bétournay M.C., 2009, Abandoned metal mine stability risk evaluation. Risk Anal. 29 (10): pp. 1355-1370. Castellanza R., Lollino P., Ciantia M.O., 2018, A methodological approach to assess the hazard of underground cavities subjected to environmental weathering. Tunnel Underground Space Technology, 82, pp. 278 292. De Silva F., Lusi T., Ruotolo M, Flora A., Ramondini M., Urciuoli G., 2022, A simplified approach to assess the stability of tuff cavities accounting for the spatial variability of the shear strength and the presence of joints. Geotechnical Engng for the Preservation of Monuments and Historic Sites III. Lancellotta, Viggiani, Flora, de Silva & mele (Eds), Taylor and Francis (London). Fazio N.L., Perrotti M., Lollino P., Parise M., Vattano M., Madonia G., Di Maggio C., 2017, A three-dimensional back analysis of the collapse of an underground cavity in soft rocks. Engineering Geology, vol. 238, pp. 301-311. Ferrero A.M., Segalini A., Giani G.P., 2010, Stability analysis of historic underground quarries. Comput. Geotech. 37: pp. 476-486. Fiore A., Fazio N.L., Lollino P., Luisi M., Miccoli N.M., Pagliarulo R., Perrotti M., Pisano L., Spalluto L., Vennari C., Vessia G., Parise M., 2018, Evaluating the susceptibility to anthropogenic sinkholes in Apulian calcarenites, southern Italy. In: Parise M, Gabrovsek F, Kaufmann G, Ravbar N (Eds.), Advances in Karst Research: Theory, Fieldwork and Applications. Geological Society, London, Special Publications, 466, pp. 381-396. Gutiérrez F., Parise, M., De Waele, J., Jourde, H., 2014, A review on natural and human-induced geohazards and impacts in karst. Earth Science Reviews 138, pp. 61-88. Hammah R.E., Yacoub T., Corkum B. & Curran J.H., 2008, The Practical Modelling of Discontinuous Rock Masses with Finite Element Analysis. 42nd US Rock Mechanics Symposium, San Francisco (USA), June 29-July 2, 2008. Li L.H., Yang Z.F., Yue Z.Q., Zhang L.Q., 2009, Engineering geological characteristics, failure modes and protective measures of Longyou rock caverns of 2000 years old. Tunneling and Underground Space Technology 24: pp. 190-207. Lollino P. & Andriani G.F., 2017, Role of brittle behaviour of soft calcarenites under low confinement: laborator observations and numerical investigation. Rock Mechanics & Rock Engineering, 50(7), 1863-1882. DOI: 10.1007/s00603-017-1188-0. Lollino P. & Parise M., 2023, Sinkhole hazard quantitative assessment: Insights from the application of numerical modelling techniques. In: Land L., Kromhout C. & Suter S. (Eds.), Proceedings of the 17th Multidisciplinary Conference on Sinkholes and the Engineering and Environmental Impacts of Karst, Tampa (Florida, USA), 27-31 March 2023, NCKRI Symposium no. 9, pp. 141-150. Lollino P., Parise M. & Reina A., 2004, Numerical analysis of the behaviour of a karst cavern at Castellana-Grotte, Italy. In Numerical Modeling of discrete materials in Geotechnical Engineering, Civil Engineering & Earth Science, Konietzky Ed., A.A.Balkema Publishers, pp. 49-55. Munjiza A., 2004, The combined finite–discrete element method. Wiley, Hoboken. Parise M., 2010, The impacts of quarrying in the Apulian karst. In: Carrasco, F., La Moreaux, J.W., Duran Valsero, J.J., Andreo, B. (Eds.), Advances in Research in Karst Media. Springer, pp. 441-447. Parise M., 2012, A present risk from past activities: Sinkhole occurrence above underground quarries. Carbonates and Evaporites 27 (2), pp. 109-118. Parise M., 2015, A procedure for evaluating the susceptibility to natural and anthropogenic sinkholes. Georisk 9 (4): pp. 272-285. Parise M., 2019, Sinkholes. In: White, W.B., Culver, D.C., Pipan, T. (Eds.), Encyclopedia of Caves, 3rd edn. Academic Press, Elsevier, pp. 934-942. Parise M., 2022, Sinkholes, Subsidence and Related Mass Movements. In: Shroder J.J.F. (Ed.), Treatise on Geomorphology, vol. 5. Elsevier, Academic Press, pp. 200-220. https://dx.doi.org/10.1016/B978-0-12-818234- 5.00029-8. ISBN: 9780128182345. Parise M., Lollino P., 2011, A preliminary analysis of failure mechanisms in karst and man-made underground caves in Southern Italy. Geomorphology, vol. 134 (1-2), pp. 132-143. Perrotti M., Lollino P., Fazio N.L., Pisano L., Vessia G., Parise M., Fiore A., Luisi M., 2018, Finite element-based stability charts for underground cavities in soft calcarenites. Int. J. Geomech., 18, 7. Van Den Eeckhaut M., Poesen J., Dusar M., Martens V., Duchateau P., 2007, Sinkhole formation above underground limestone quarries: A case study in South Limburg (Belgium). Geomorphology 91(1): pp. 19-37. Vattano M., Di Maggio C., Madonia G., Parise M., Lollino P., Bonamini M., 2013, Examples of anthropogenic sinkholes in Sicily and comparison with similar phenomena in southern Italy. Proc. 13th Multidisc. Conf., 6–10 May 2013, Carlsbad, New Mexico, NCKRI Symposium, vol. 2, pp. 263-271. Zhang Q.B., He L., Zhu W.S., 2016, Displacement measurement techniques and numerical verification in 3D geomechanical model tests of an underground cavern group. Tunnelling and Underground Space Technology 56: pp. 54-64.

Ancient underground hydraulic works

Roberto Bixio, 2003 Genoa “Window” on the sixteenth-century Bastion of Acquasola (Genoa, Italy), now completely buried under the homonymous city public park. Icon of the full original painting. (Watercolour, 17×13 cm)

85 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   OBRUK Cave Research Group, Açıkhava Apt. 16/7 Nişantaşı, Istanbul, Turkey - [email protected] Ahmet Çelebi Qastel of Gaziantep (Turkey) Ali Yamaç1 Abstract It has long been known that Gaziantep is home to numerous limestone-excavated underground structures. Some of the subterranean spaces are now yarn ateliers, while others were originally used as cisterns or storage areas. Despite being a big and intricate system, underground water structures are another crucial aspect of our survey. This water supply system consists of a network of canals cut into limestone rock called “livas” in the area. It is similar with the “qanat” or “karez” system, which was invented in Persia between 2000 and 3000 years ago and is still in use in many countries today. On the other hand, another underground structure built for public usage and locally known as “qastel” is a unique structure of Gaziantep. With their washing places, pools, and toilets, six remaining “qastels” are the only examples of public underground water usage. As OBRUK Cave Research Group, we started the “Gaziantep Underground Structures Inventory Project”, in 2012. This project involves researching, surveying, mapping, and documenting all the underground and rock-cut structures found in Kayseri province. It is carried out in accordance with a triple protocol with the Foundation for the Protection and Promotion of the Environment and Cultural Heritage (ÇEKÜL) and Kayseri Metropolitan Municipality. These underground water structures were explored and surveyed within the scope of this project. In this article, the research of one of these six qastels, Ahmet Çelebi Qastel, and the problems in the restoration process are explained. Keywords: Gaziantep, livas, qanat, karez, qastel, Ahmet Çelebi Qastel. Introduction As OBRUK Cave Research Group, we started the “Gaziantep Underground Structures Inventory Project” in 2012. This ongoing project is carried out with the ÇEKÜL Foundation and the Gaziantep Metropolitan Municipality, with a tripartite protocol and the primary purpose is to survey and inventory all underground structures in and around Gaziantep. Gaziantep, located in southeastern Turkey (fig. 1), has been continuously inhabited since 3000 BC. The former location of the city was at Düllük Village, 9 kilometers from the city center, according to archeological excavations. After this Paleolithic-era village was abandoned, the city developed around the castle in the eleventh century. The plateau that houses the city of Gaziantep and its periphery mainly consists of soft Upper-Middle Eocene-aged limestone and chalk (MTA, 1997). This limestone formation, which is relatively easy to carve into, has determined the city’s development and character in many respects. Gaziantep has hundreds of underground structures carved into this rock formation. Some underground structures were used as storage facilities or cisterns, while others are used as yarn ateliers today. The absence of a water table was Gaziantep’s biggest issue. The city, which was founded on a hill and expanded through time, overcame this difficulty by using underground aqueducts to bring water from neighbouring springs. Tens of kilometers of local underground aqueducts, called “livas” were dug under the city situated over this easy-to-dig soft limestone. On the other hand, apart from this water distribution system with similar examples throughout the world, Fig. 1 – Location map showing Gaziantep (from Yamaç & Okuducu, 2017). Fig. 2 – Sketch profile of Gaziantep showing the underground aqueducts (livas), qastels and wells (drawing A. Yamaç).

86 Ahmet Çelebi Qastel of Gaziantep (Turkey) there are impressive underground public areas, locally called “qastel,” unique to Gaziantep, where the water coming through these underground channels is used publicly (fig. 2). Modern water distribution reached the city of Gaziantep only in the 1950s. Unfortunately, two kinds of destruction and neglect have brought this ancient livas system that served the city for centuries near destruction. One of these two factors is individual users filling the wells they do not need anymore and, even worse, filling them with mortar. The other is that, the modern municipalities damage the system beyond repair during infrastructure works for electricity, natural gas, water, etc.. During this project we explored and surveyed 46 underground structures in and around Gaziantep city. Within the same project, ancient water system; livas and qastels of Gaziantep has been meticulously surveyed and inventoried (Uçar 2016; Yamaç and Okuducu, 2017; Uçar 2018). As representing a unique component of the cultural heritage, these structures have been considered for the candidacy of Unesco World Heritage. Our support to protect and restore all these underground structures, which were accepted to the UNESCO World Heritage Site Tentative List in 2018, continues. There are numerous water resources around the plateau upon which Gaziantep is located. Despite all these resources, there is no water table under the city. Therefore, long tunnels were dug to bring water to Gaziantep from those springs. The length of some of these underground aqueducts reaches a few kilometers. They are the very image of ancient “qanat” or “karez” that were first implemented 2000–3000 years ago in Iran and encountered in many countries such as Morocco, Persia, Algeria, Egypt, and China, and Italy (Mazloum, 1936; Lightfoot, 1996, 1997, 2000; Wessels, 2000; Castellani, 2001; Yazdi and Khaneiki, 2012, 2019). Although the beginning of the digging for this aqueduct system cannot be determined, it is known that the system was in service until the middle of the 20th century. The main components of the system are underground passageways with well-calculated, smooth slopes that convey water from a primary source relatively outside the town to large settlement areas and distribute it through wells created where water was required to the city center. Another problem was meeting the livas below from the wells to be dug at the surface. It is difficult today to determine how this was achieved. The landscape in downtown Gaziantep is very rugged, with significant differences between the levels of neighborhoods or even houses in the same neighborhood. Making a well dug from the garden of a house meet with an active livas through a precisely calculated shaft and dug from the correct level in the correct direction is not a small feat. We could determine that the workflow was in this direction, i.e., from the well to the livas, in at least one sample. On the other hand, digging wells in the higher parts of the city are challenging. For example, the American Hospital’s well that leads to the livas is 44 meters deep. Apart from its scope, the engineering calculations for the system are also impressive for the period. It is obvious that obtaining the correct slope at every point to ensure a constant flow in the channels for tens of kilometers was very important in terms of the performance of the system. The scope of the labor may be better understood by considering that the work was conducted underground, in the dark, in harsh conditions, and with limited technological means. The dimensions of this system are imposing. The primary water source that fed the old settlement was situated at the current location of Alleben Pond. Although the distance from Alleben to the city center is around 10 km, the total distance of the interconnected livas system would reach hundreds of kilometers. Various aqueducts like those in the Gaziantep War Museum Cave, Tutun Han Cave, Nadir Bey House, and Omer Ersoy Culture Center are all connected to the same underground water system. On the other hand, it has been determined that many wells investigated during the project are also connected to this system. Unfortunately, it has not been possible to follow this underground aqueduct system for long distances in Gaziantep, which has been severely damaged by modern construction over the last 40–50 years. On the other hand, 1.980 m have been measured so far in Pancarlı Livas, which starts from a primary water source 17 km west of the city (fig. 3). Fig. 3 – With a total length of nearly 2 km, Pancarlı Livas is the longest underground aqueduct surveyed in Gaziantep so far (photo A. E. Keskin).

87 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Public Underground Water Structures of Gaziantep: “Qastel” Public fountains offered for everyday use were indispensable elements of every settlement. There are countless fountains in the squares of almost every city and every village, from Europe to Asia. These public fountains met the citizens’ water needs and provided social interaction, thus serving a sociological purpose. What if you do not have the groundwater to build a fountain? The people of Gaziantep answered this question with an ingenious solution hundreds of years ago. In addition to the underground aqueducts, or livas, structures specific to Gaziantep known as “qastel” are another outstanding aspect of the city’s underground water system. These buildings are communal areas built to provide access to the clean water in the city for a larger audience, which makes them different from the individual wells. These buildings were designed to provide convenient access to clean water for washing and use, as well as to function as gathering places for people (Çam, 1982, 2006). Almost all qastels have washrooms, pools for washing clothes, and even public toilets. Inside the structure, clean and dirty water from the underground aqueducts are separated and flowed in different directions. Although there were at least 15 qastels in Gaziantep until 60 years ago, only six examples remain today. Others, unfortunately, have disappeared under modern constructions or renovations. The sad thing is that the elders of Gaziantep still remember the locations and names of these vanished qastels. Today, these six qastels, namely; Ahmet Çelebi, Ihsaniye, Imam-ı Gazali, Kozluca, Pişirici, and Şeyh Fethullah, have all been restored. Although these structures seem religious, because two of the six samples preserved until today are next to mosques, some findings suggest that the qastel structures date farther back in time. It may be thought that, in many cases, the mosques built later took over the existing qastel structures. Moreover, the fact that the two qastels have small underground praying places (masjids) suggests that they were built before the mosques above. The best example is the Pişirici Qastel, built in 1283, which is one of the oldest structures in the city (fig. 4). This structure can be considered an engineering masterpiece spanning hundreds of years, with its system still working today. Despite all the destruction in the underground aqueducts, clean water still flows to Pişirici Qastel; it divides into several branches inside the building and leads to different pools, while the dirty water line passes under the toilets and goes in a completely different direction. Fig. 4 – Dated 1283, Pişirici Qastel is one of the oldest structures in Gaziantep (from Yamaç & Okuducu, 2017; photo A.E. Keskin).

88 Ahmet Çelebi Qastel of Gaziantep (Turkey) Ahmet Çelebi Qastel Ahmet Çelebi Qastel does not have an inscription, and its date of construction is unknown. However, it was inevitably constructed before the mosque on the site, which bears the same name and is dated 1672. Namely, to the west of the Ahmet Çelebi Mosque, above the qastel, is an open space with a deep niche. With this location, this open place reminds of a masjid, and there will be no need for a masjid when there is a mosque next to it. For this reason, this structure must have been built before the mosque. Although there are masjids inside the Pişirici and İhsaniye qastels, Ahmet Çelebi Qastel’s masjid was most likely built on top due to a lack of underground space. Therefore, the qastel and the upper masjid must be older than the mosque (Çam, 2006). Qastel is 9.50 m below the surface and can be reached by 49 steps of stairs. The floor is approximately 5 x 8 m in size, and there is a 2 x 2 m pool in this area close to the eastern staircase. The top of the Qastel has a natural rock that is 2.5–3 m thick, and a 1 x 1 m lighting hole has been dug near its western end (fig. 5). The restoration of this building, which is architecturally unlike any other four qastels except for Imam-i Gazali, started last year. In fact, the problems that could arise even before the restoration began were obvious. There were five different livas’ in this structure (fig. 6). True, all other qastels also had livas’, as it should be, but the livas’ of these other qastels were either not working or were blocked and short. Ahmet Çelebi Qastel’s five different livas’ were much longer than the others and more active. It was not possible to understand which livas were initially filling and which were discharging because they were all blocked after a point. As if all these problems were not enough, the original floor was covered with stone during a planless and thoughtless restoration about thirty years ago, and the entrances of L4 and L5 livas’ were partially covered (fig. 6). When we removed the floor covering applied in the last restoration, we were planning to find the original state of the toilets, which are indispensable for a qastel, and indeed, we found the toilets. On the other hand, when the floor covering was removed, a very different picture and much more serious problems emerged. We suddenly realized that the entire original Qastel was carved into the bedrock, and under the floor covering, it was seen that there was another, original pool carved into the bedrock. Probably because the original pool could not be used, a new one was built on top of it. Apart from this, there were traces of structures in the bedrock, some of which still need to be understood. All these problems can be seen in the restitution plan (fig. 7). Another problem was the discovery of another livas under the L3 livas and parallel to it. Before we could fully decipher the function of the other five, a sixth livas was found (fig. 8). Fig. 6 – General plan of Ahmet Çelebi Qastel with the livas’ (modified from Yamaç & Okuducu, 2017). Fig. 5 – Ahmet Çelebi Qastel before the restoration (photo A. Yamaç).

89 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Ultimately, it was decided to refill the opened floor and restore only the exposed toilets. Also, it was decided to discharge the water coming from L1 and L2 in the direction of L3 after passing the pool. This was probably not the original system, but decades of indifference and destruction led us to this palliative solution today. Conclusion Ahmet Çelebi Qastel is neither the largest nor architecturally the most important of the six qastels in Gaziantep today. Nevertheless, this building, built 400 years ago, is an excellent example of the cost and effort spent to meet a simple social need. On the other hand, the aqueducts under the city, some of which are still active, can no longer be explored as a result of years of destruction, and therefore we do not know the functions of the five different livas’ surrounding Ahmet Çelebi Qastel today and how we will restore this impressive system. This building, which has been functioning flawlessly for centuries and which we cannot even restore today, is another proof of how alienated we have been from our historical values over the years. Bibliography Castellani V., 2001, Acqua, acquedotti e qanat. Opera Ipogea 2: pp. 25-32. Çam N., 1982, Gaziantep’te “Kastel” Adı Verilen Su Tesisleri. In, Milletlerarası Türkoloji Kongresi (International Turkology Congress), İstanbul. Çam N., 2006, Türk Kültür Varlıkları Envanteri: Gaziantep (The Inventory of Turkish Cultural Proporties: Gaziantep), Ankara. Lightfoot D., 1996, Syrian qanat Romani: history, ecology, abandonment. Journal of Arid Environments 33: pp. 321-336. Fig. 7 – Restitution plan of Ahmet Çelebi Qastel (drawing S. Savcılı). Fig. 8 – Survey in the L3 livas of Ahmet Çelebi Qastel (photo A. E. Keskin).

90 Ahmet Çelebi Qastel of Gaziantep (Turkey) Lightfoot D., 1997, Qanats in the Levant: Hydraulic Technology at the Periphery of Early Empires. Technology and Culture 38: pp. 432-451. Lightfoot D., 2000, The Origin and Diffusion of Qanats in Arabia: New Evidence from the Northern and Southern Peninsula. The Geographical Journal 166: pp. 215-226. Mazloum S., 1936, L’ancienne Canalisation d’eau d’Alep. Documents d’Etudes Orientales de L’Institut Français de Damas. MTA Maden Tetkik ve Arama Genel Müdürlüğü (General Directorate of Mineral Research and Exploration), 1997, Geological Map of Gaziantep - K24 Quadrangle, Ankara. Uçar M., 2016, Gaziantep Tarihi Su Sistemi ve Su Yapıları. METU JFA 33: pp. 73-100. Uçar M., 2018, Gaziantep Livasları, Kastelleri ve Yer Altı Su Yapıları, Gaziantep. Wessels K., 2000, Renovating Qanats in a changing world, a case study in Syria. International Symposium on Qanats, 2000, Yazd, Iran. Yamaç A., Okuducu Z., 2017, Underground Hydraulic Structures of Gaziantep (Southeast Turkey): Livas and Qastels. M. Parise, C. Galeazzi, R. Bixio and A. Yamaç (eds.), HYPOGEA, 2017, Cappadocia, Turkey. Yazdi A., Khaneiki M., 2012, Qanat in its cradle, Shahandeh Publication. Yazdi A., Khaneiki M., 2019, Veins of the Desert: A Review on Qanat / Falaj / Karez, River Publishers.

91 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   University of Pavia, Dipartimento di Studi Umanistici, Piazza del Lino 2, Pavia *  Reference author: [email protected] Water monuments in Hittite and Neo-Hittite periods: structure, functions, and connection with the “other world” Maria Elena Balza1 , Marco Capardoni1 , Clelia Mora1,* Abstract The great Hittite kingdom ruled over Anatolia and the northern area of Syria for a few centuries during the 2nd millennium BCE. After the end of this kingdom (12th century BCE), and after a period during which written documentation was very limited (12-10th centuries BCE), new written and archaeological sources show a different situation, fragmented into small principalities located in Anatolian areas east and south of the heartland of the previous Hittite kingdom. The present paper will focus on two important and impressive Hittite and Neo-Hittite monuments dating back to the 13th (Eflatunpınar) and 8th centuries BCE (İvriz) that are connected to water sources and/or intended to collect water. We will first examine their characteristics and environmental location, as far as possible according to archaeological and historical investigations; we will then attempt to explain, on the basis of archaeological and textual data and of previous studies, their practical function and their connections with cult and religious aspects, if any. These monuments and structures also show the peculiarities of the Anatolian territory compared to other regions of the Near East as regards the water supply. Keywords: Ancient Anatolia, Hittite and Neo-Hittite Water Monuments, Eflatunpınar, İvriz. Introduction Partly following the path traced by Wittfogel (1957; cf. also the discussion in Vidal-Naquet, 1964), historical, political, and social studies on ancient Near Eastern settlements and societies often dealt with water policy and with the production of many water-based historical models. They mainly focused on the geomorphological features of Mesopotamia and Egypt, and on the political and social effects of their agricultural production based on extensive irrigation systems. Unlike these areas, in a similar chronological and cultural context, the Anatolian territory was characterized by a more widespread water supply, provided by various types of sources. The great Hittite kingdom ruled over Anatolia and the northern area of Syria for about five centuries, during the 2nd millennium BCE. The Hittite kings of the later period (13th-early 12th centuries BCE) left us a series of works and monuments, which were often impressive, connected to springs or water basins. These structures are in many cases enriched by inscriptions and images, mainly in relief on stone, referring to the king and tutelary deities. In addition to the traditional cuneiform script, an important cultural aspect to consider is that the Hittite administration had used a new script, called ‘Anatolian Hieroglyphic (AH) writing’, since at least the 14th century BCE. This script was also used for inscriptions placed near monuments or structures connected to water. In the ancient common belief, the water that gushes and flows from the karst terrains was connected to the otherworldly realm (Harmanşah, 2019: 2211 ) and the king, present with his image, also derived prestige from it. Some clues and above all specific geo-archaeological evidence seem to testify that the Hittite ‘hydraulic’ works had primarily practical purposes, to support agriculture and livestock, especially in areas that remained marginal with respect to the main centers of power (among others, see Emre, 1993; Harmanşah, 2018). In the final phase of the late Bronze Age, moreover, some written and archaeological evidence seems to suggest one or more periods of crisis related to changing climatic conditions and water shortages in the eastern Mediterranean area. At this stage, the water problem in Anatolia seems, therefore, to have become even more pressing (Divon, 2008; De Martino, 2018 with previous bibliography). In Hittite cuneiform texts some words seem to re1 “Emerging from the orifices of Anatolian karst geologies, water is believed to connect the divine Underworld and the surface of the earth through caves, sinkholes, springs, and river sources”.

92 Water monuments in Hittite and Neo-Hittite periods: structure, functions, and connection with the “other world” fer to peculiarities of the Anatolian karst landscape, such as sinkholes, cave mouths, and swallow holes (Mora et al., 2017). One of the most interesting terms is DKASKAL.KUR, a composition of two ideograms meaning, respectively, ‘way, passage’ (KASKAL) and ‘territory, country’ (KUR), preceded by the determinative sign ‘god’ (DINGIR): ‘divine passage/opening into terrain’, generally translated ‘underground watercourse’ or (Hawkins, 2000: 293) ‘karstic slot, pot-hole’. This term has also been connected by Hawkins (1995, 2015) to a formulaic expression occurring in some inscriptions in AH writing: (DEUS) VIA+TERRA (translated ‘divine earthroad’), but some scholars have expressed doubts about this interpretation (cf. discussion in Erbil and Mouton, 2012: 59-60; cf. also Payne, 2019: fn. 49, 259). At the beginning of the 12th century BCE, the Hittite reigning king and the court abandoned the capital city for a destination that is still unknown to us. After a period of a few centuries, during which written documentation was very limited – making historical reconstruction of the 12-10th centuries BCE extremely difficult – new written and archaeological sources show a completely different socio-political situation, fragmented into small principalities located in Anatolian areas east and south of the heartland of the previous Hittite kingdom. These new states, called Neo-Hittites, displayed documentation in AH writing and no longer used the traditional cuneiform script, at least according to what is preserved. Some important monuments connected to water sources, located in the south-eastern areas of Anatolia, also date back to this period of ‘rebirth’. We will focus below on two of the main Hittite and NeoHittite monumental works related to underground and spring waters, respectively the Eflatunpınar basin (13th century BCE) and the İvriz reliefs (8th century BCE) (fig. 1). The spring of Eflatunpınar and the Hittite pool The remains of at least three facilities linked to underground and stream water provide evidence (cf. their AH inscriptions) of king Tuthaliya IV’s attempt to regiment the waters (second half of the 13th century BCE). These remains are (i) the structure of Yalburt yaylası, a rectangular shaped pool in the Konya plain, in the vicinity of Ilgın; and (ii) the large stone blocks of Karakuyu (Pınarbaşı, Kayseri; Emre, 1993) and (iii) Fig. 1 – Map of Central-South Anatolia with the main sites mentioned in the text (map by M. Capardoni).

93 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa the rectangular stone block of Köylütolu, discovered in a plain on the Ilgın-Kadınhanı road near Köylütolu, both belonging to a dam (Ehringhaus, 2005: 37 ff. with previous literature). In addition to these works, there are at least two other structures that can be attributed to the same chronological phase, the Alacahöyük/Gölpınar dam and the Eflatunpınar pool (Ehringhaus, 2005: 50-56 ff.). Among these structures, the most impressive is undoubtedly the Eflatunpınar pool (Kohlmeyer, 1983: 34-43; Rossner, 1988: 67-74; Börkher-Klähn, 1993; Emre, 2002; Bachmann and Özenir, 2004; Ehringhaus, 2005: 50-57). We would therefore like to focus on the description of this monument connected with underground water. Although no inscription has been found on the monument, its style seems to point to the final phase of the 13th century BCE, probably the time of Tuthaliya IV (Hawkins, 2015: 2). The spring Eflatunpınar lies in the Beyşehir district, west of Konya, 6 km north-east of Beyşehir Lake, wherein a series of springs gushes forth and produces a stream flowing into Beyşehir Lake. This perennial water source is embedded in an artificial complex, the most spectacular part of which is represented by a façade protruding from water and entirely covered with reliefs (figs. 2 and 3). In the center of the scene, there are two seated figures (a male and a female), each of which is surmounted by the representation of a winged sun. These two main figures are surrounded by hybrid beings, whose function is to carry the wings of the two suns, as well as the wings of an even larger winged sun that covers the entire representation. Above the larger upper winged sun, another one presumably of the same length is missing. At the bottom of the scene, under the feet of the two seated figures, there are five other partially visible figures, probably mountaingods. Three of these figures are characterized by the presence of openings in their bodies, through which the water of the spring was supposed to flow, thus creating a rather spectacular theatrical effect. The two seated figures in the center of the scene have been interpreted in different ways, but most interpretations highlight their divine nature. Scholars have identified them as the proto-Hattian solar couple, as the Storm God and the Sun Goddess of the city of Arinna, or as the Sun God of the Sky and the Sun Goddess of the Earth (Bittel, 1953: 4-5; Börker-Klähn and Börker, 1975: 34 ff.; Kohlmeyer, 1983: 42-43). Based on these interpretations, the scene that unfolds before the eyes of the audience could be interpreted as a cosmological representation of the world, with the Sun at the top, the Earth with its mountains and springs at the bottom, and the gods in the middle functioning as a link between the different elements that constitute the world (Bittel, 1953; Orthmann, 1964; Kohlmeyer, 1983; Ehringhaus, 2005: 50ff.; Erbil and Mouton, 2012; Bachmann, 2017). The representation of the winged sun, however, also recalls Hittite kingship. The winged sun, indeed, represents one of the symbols of Hittite royalty (at least since the 14th century BCE). In this regard, Erbil and Mouton (2012: 70) suggested that “the deities figuring on the monument seem to be the tutelary gods of the Hittite king himself.” But another hypothesis was put forward by Harmanşah (2015), according to whom the monument could be seen as an attempt to reproduce the appearance of a mountain spring, gushing out of the natural stone, at a place that is far from the mountains, in the middle of a valley. Within the Hittite belief system, in fact, springs, like many other natural elements, were considered places of great importance: “the Fig. 2 – The Eflatunpınar monument (courtesy of T. Bilgin).

94 Water monuments in Hittite and Neo-Hittite periods: structure, functions, and connection with the “other world” greater part of the landscape of Hittite Anatolia was in some sense sacred, in that the mountains, rivers and springs were so regarded (…).” (Hawkins, 2015: 1). The Eflatunpınar monument may therefore be a mimetic work that recreates a mountain spring. The basin is indeed made of the same volcanic stone of the mountains of the Anatolian plateau, and the water from the spring is channeled to gush out of the holes dug in the statues that decorate the basin just as it would have done in a real mountain spring (Harmanşah, 2015: 79). The Eflatunpınar basin may then have been both a structure that was used for groundwater collection and storage (so important for the survival of humans and animals) and a monument that depended on the sanctity of the spring around which it was erected. In this sense, it might recall the divine openings on the Earth capable of connecting the world of humans with the Other World (see above). One may also wonder whether Eflatunpınar could be interpreted as a place of political or religious performances. Regarding this possibility, it has been suggested that ‘this sacred pool was an important station for the pilgrimage of the Great king during cultic festivals’ (Erbil and Mouton, 2012: 70). And this hypothesis may be partly confirmed by the presence of a settlement dating back to the 2nd millennium BCE in the vicinity of Eflatunpınar (Özenir, 2001: 540; cf. also Erbil 2019). This and other Hittite water basins could therefore have served as places where, not only through great works of engineering, but also through prayers and rituals, humans tried to prevent underground springs from drying up definitively, drought and death from taking over the land, and the passage between the world of human beings and gods from being closed off. The rock reliefs of İvriz The rock reliefs of İvriz (named after the nearby village, today renamed as Aydınkent) are located at the feet of the northern slopes of the Bolkar mountains in the district of Halkapınar (Konya Province). The local, karstic landscape is characterized by springs and sinkholes. The water flowing from these springs feeds the river İvriz Çayı that represents one of the main water sources of the valley down to Ereğli (fig. 1). The water gush of the springs is very abundant during spring and dries up during the summer (Ehringhaus, 2014: 48-50). The large rock relief İvriz 1 appears spatially associated with a group of springs on the edge of the river which originates a few hundred meters Fig. 3 – The façade of the Eflatunpınar monument (courtesy of T. Bilgin).

95 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa upstream, from a spring on the hillside. The rock relief İvriz 2 is located just a few meters from the latter spring (fig. 4). Another rock relief is found in a narrow, steep valley known as Ambar Deresi, c.1km north of the village of Aydınkent. İvriz 1 (fig. 5) depicts two human figures facing each other at different scales. Thanks to two AH inscriptions in relief accompanying the figures, it is possible to identify the larger figure on the left as the god Tarhunza (h. c.4.2m) and the smaller one on the right as the king Warpalawa of Tuwana (h. c.2.5m). The latter commissioned the monument as well, as revealed by a third AH inscription mentioning the name of the scribe and qualifying him as servant of Warpalawa (Hawkins, 2000: 516-518). The Storm-god wears a tunic with short sleeves and two volutes at the hem, one belt supporting a sickle (Şahin, 1999), a horned headdress, boots with upturned toes and bracelets (Aro, 2003: 335-336). He is represented grabbing grapes and ears of grain stemming from the ground with his hands. Similar representations of the Storm-god are also attested from elsewhere in the area of ancient city of Tuwana (classical Tyana, today Kemerhisar, south of Niğde). A strict parallel to it is also provided by the text of the Sultanhan stele, which the local ruler, Sarwatiwaras, dedicated to Tarhunza of the Vineyard, who ‘came […] and the corn-stems burgeoned forth at his foot and the vine was good here’ (Hawkins, 2000: 463-472). This god was associated with agricultural activities and the fertility of the land. The king is represented in front of the god with the hands joined in an adoration gesture. He wears a richly decorated long vest with a belt and a cloak, possibly indigenous in origin (Aro, 2003: 336), fastened by a Phrygian brooch (Ehringhaus, 2014: 51), his head covered by a hat with a tuft on front. Fig. 4 – The location of the rock reliefs İvriz 1 and İvriz 2 (from Ehringhaus, 2014: fig. 49, p. 49; modified by the authors). Fig. 5 – The rock relief İvriz 1 (photo by C. Mora).

96 Water monuments in Hittite and Neo-Hittite periods: structure, functions, and connection with the “other world” Scholars have generally considered the stylistic renditions of the curled hair and beards of the god and the king as of Assyrian influence (lastly Aro, 2003, and Ehringhaus, 2014). Warpalawa was the king of a Neo-Hittite state centered on its capital Tuwana, which extended across southern Cappadocia. He is known from local inscriptions and monuments realised by himself or by other prominent figures. It is generally accepted to identify the king Warpalawa of Tuwana attested in local sources with Urballa of Tuhana mentioned in Assyrian sources between 738 and 710-709 BCE. The reign of Warpalawa is therefore assigned to the second half of the 8th century BCE. In this period, the state of Tuwana probably represented one of the major polities among the Tabalian states. These states held the status of client states of the Assyrian power, however, the relationships with the central power were complicated by the international competition over the region, which involved also two other major players, namely the Phrygians and the Urartians (Melville, 2010). The İvriz 1 relief with its multiple influences well reflects the international dimension of the ruling class of a Tabalian state of the time. Dating to the reign of Warpalawa is also a fragment of stele depicting the lower part of the Storm-god in a similar fashion to the large rock relief, accompanied by Phoenician and AH inscriptions mentioning the king, the erection of the stele to Tarhunza and list of offerings (Dinçol, 1994; Hawkins, 2000: 526). This stele was recovered in 1986 c.75m upstream from the rock relief along with a giant head (h. 0.7m) belonging to a colossal statue, possibly portraying Warpalawa (Dinçol, 1994). These findings would attest the existence of a major cultic complex dedicated to Tarhunza in the area during the second half of the 8th century BCE. The poorly preserved rock relief İvriz 2 (fig. 6) represents a fragmentary offering scene, associated to carved steps leading to a rectangular pit carved in the rock interpreted as a stele basement (Bier 1976; D’Alfonso, 2020 with previous bibliography). Its dating is debated, spanning from earlier than the end of the 9th century to the end of the 8th century BCE (D’Alfonso, 2020 for a discussion of the topic). If the dating of İvriz 2 to a period earlier than Warpalawa reign holds true, it would point out the long-lasting attractiveness and supposedly holy character of the place well beyond the limits of Warpalawa reign, whose monumental rendition of the place has thus to be considered as a major event in a long-lasting tradition. A similar long-lasting appeal has been proposed also for the Ambar Deresi valley (Rojas and Sergueenkova, 2014: 147-149). Close to the head of this valley, a large rock relief with the same adoration scene of İvriz 1, but without inscriptions, was carved on a steep cliff. Close to this second rock relief of Warpalawa, Christian monks set up a community during the Byzantine period including different buildings and a church with frescoes. What matters on this occasion is that also for the Ambar Deresi relief an association with ephemeral and subterranean waters can be proposed. On the one hand, the Turkish term ‘dere’, indicates a valley with an intermittent, seasonal stream (Ehringhaus, 2014: 49-50). On the other hand, in one cave located few meters away from the relief, the sound of flowing water was reported during September 2015 but not in June 2016 (Maner, 2017: 104). Pottery fragments from the cave would point out some activities during the Iron Age, whilst a circular pit carved in the rock at the entrance of the grotto is interpreted as a libation hole (Maner, 2017: 104). Based mainly on written sources from the 2nd millennium BCE, scholars have proposed that the drying up of water springing from the underground cavities of İvriz may have been interpreted as a tangible manifestation of the absence of the god Tarhunza. The main aim of the cult developed around the springs, therefore, would be the propitiation Fig. 6 – The rock relief İvriz 2 (from Bier, 1976: fig. 5, p. 120).

97 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa of the return of the Weather-god (Bier, 1976: 124- 126; Ehringhaus, 2014: 60-61). During his reign, Warpalawa might have claimed such an important event as a royal prerogative, thus giving it also a political dimension. The archeological sites of İvriz and Ambar Deresi thus provide examples of how the natural phenomenon of the water cycle in a karstic landscape characterized by underground passages was conceptualized with religious, ideological, and political implications which went far beyond the sole satisfaction of practical needs. Bibliography Aro S., 2003, Art and Architecture, in Melchert H.C. (ed), The Luwians: pp. 281-337. Brill, Leiden - Boston. Bachmann M., 2017, Manifestation göttlicher Präsenz. Das Quellheiligtum Eflatun Pınar, in Schachner A.(ed), Innovation versus Beharrung. Was macht den Unterschied des hethitischen Reichs im Anatolien des 2. Jahrtausends v. Chr.? Internationaler Workshop zu Ehren von J. Seeher Istanbul, 23-24. Mai 2014: pp. 105-120. Ege Yayinlari, İstanbul. Bachmann M., Özenir, S. 2004, Das Quellheiligtum Eflatun Pınar, Archäologischer Anzeiger 2004/1: pp. 85-122. DAI, Berlin. Bier L., 1976, A Second Hittite Relief at Ivriz, Journal of Near Eastern Studies 35/2: pp. 115-126. University of Chicago Press. Bittel K., 1953, Beitrag zu Eflâtun-Pinar, Bibliotheca Orientalis 10: pp. 2-5. NINO, Leiden. Börker-Klähn J., 1993, Noch einmal Iflatun Pinar, in Mellink M.J., Porada E., Özgüç T. (eds), Aspects of Art and Iconography. Anatolia and its Neighbors. Studies in Honor of N. Özgüç: pp. 339-355. Türk Tarih Kurumu Basımevi, Ankara. Börker-Klähn J., Börker, Ch., 1975, Eflatun Pınar. Zur Rekonstruktion, Deutung und Datierung, Jahrbuch des Deutschen Archäologischen Institut 90: pp. 1-41. DAI, Berlin. D’Alfonso L., 2020, Reorganization vs. Resilience in Early Iron Age Monumental Art of Central Anatolia, in Cammarosano M., de Vecchi E., Viano M. (eds), talugaeš witteš. Ancient Near Eastern Studies Presented to S. de Martino on the Occasion of his 65th Birthday: pp. 81-101. Zaphon, Münster. De Martino S., 2018, The Fall of the Hittite Kingdom, Mesopotamia 53: pp. 23-48. CRAST, Firenze. Dinçol B., 1994, New Archaeological and Epigraphical Finds from Ivriz: A Preliminary Report, Tel Aviv 21/1: pp. 117-128. Institute of Archaeology of Tel Aviv University, London. Divon S., 2008, A Survey of the Textual Evidence for “Food Shortage” from the Late Hittite Empire, in D’Alfonso L., Cohen Y., Sürenhagen D. (eds), The City of Emar among the Late Bronze Age: pp. 101-109. Ugarit Verlag, Münster. Ehringhaus H., 2005, Götter, Herrscher, Inschriften: Die Felsreliefs der hethitischen Grossreichszeit in der Türkei. Philipp von Zabern, Mainz am Rhein, 124 pages. Ehringhaus H., 2014, Das Ende, das ein Anfang war: Felsreliefs und Felsinschriften der luwischen Staaten Kleinasiens vom 12. bis 8./7. Jahrhundert v. Chr. Nünnerich-Asmus Verlag & Media, Mainz am Rhein, 112 pages. Emre K., 1993, The Hittite Dam at Karakuyu, in Misaka T. (ed), Essays on Anatolian Archaeology: pp. 1-42. Harrassowitz, Wiesbaden. Emre K., 2002, Felsreliefs, Stelen, Orthostaten. Großplastik als monumentale Form staatlicher und religiöser Repräsentation, in Willinghöfer H. (ed), Die Hethiter und ihr Reich: pp. 218-233. Theiss, Bonn. Erbil Y., 2019, The Importance of Beyşehir in the Hittite Period, in Maner Ç. (ed), Crossroads/Kavşaklar. Konya Plain from Prehistory to the Byzantine Period/Prehistorik Çağ’dan Bizans Dönemine Konya Ovası: pp. 193-206. Ege Yayınları, İstanbul. Erbil Y., Mouton, A., 2012, Water in Ancient Anatolian Religions: An Archaeological and Philological Inquiry on the Hittite Evidence, Journal of Near Eastern Studies 71: pp. 53-74. The University of Chicago, Chicago. Harmanşah Ö., 2015, Place, Memory and Healing. An Archaeology of Anatolian Rock Monuments. Routledge, New York, 220 pages. Harmanşah Ö., 2018, Geologies of Belonging. The Political Ecology of Water in Central Anatolia, in Holt E. (ed), Water and Power in Past Societies: pp. 259-277. The State University of New York, New York. Harmanşah Ö., 2019, Cities, the Underworld, and the Infrastructure. The ecology of water in the Hittite world, in Alt S.M., Pauketat T.R. (eds), New Materialisms, Ancient Urbanisms: pp. 218-244. Routledge, London – NewYork. Hawkins J.D., 1995, The Hieroglyphic Inscription of the Sacred Pool Complex at Hattusa (SÜDBURG). Harrassowitz, Wiesbaden, 165 pages. Hawkins J.D., 2000, Corpus of Hieroglyphic Luwian Inscriptions, vol. 1: Inscriptions of the Iron Age. De Gruyter, Berlin – New York, 900 pages. Hawkins J.D., 2015, Hittite Monuments and their Sanctity, in D’Agostino A., Orsi V., Torri G. (eds), Sacred Landscapes of Hittites and Luwians: pp. 1-9. Firenze University Press, Firenze. Kohlmeyer K. 1983, Felsbilder der hethitischen Grossreichszeit, Acta Praehistorica et Archaeologica 15: pp. 7-154. Museum für Vor- und Frühgeschichte, Berlin. Maner Ç., 2017, Preliminary Report on the Fourth Season of the Konya-Ereğli Survey (KEYAR) 2016, Anatolia Antiqua 25: pp. 95-113. IFEA, İstanbul. Melville S.C., 2010, Kings of Tabal: Politics, Competition, and Conflict in a Contested Periphery, in Richardson S. (ed), Rebellions and Peripheries in the Mesopotamian World: pp. 85-107. Eisenbrauns, Winona Lake. Mora C., Balza M.E., Bixio R., De Pascale A., 2017, A Link between “Ancient Words” and the “Underground World”: Cappadocian Landscape, Rock-Cut Structures and Textual Evidence from Hittite Documentation, in Parise M., Galeazzi C., Bixio R., Yamac A. (eds), Cappadocia Hypogea 2017, Proceedings of International Congress: pp. 65-78. Ege Yayınları, İstanbul. Orthmann W., 1964, Hethitische Götterbilder, in Bittel K. et al. (eds), Vorderasiatische Archäologie. Studien und Aufsätze. A. Moortgat zum 65. Geburtstag: pp. 221-229. Mann, Berlin. Özenir S., 2001, Eflatunpınar Hitit Kutsal Anıt-Havuz 1998 Yılı Çalışmaları, in Wilhelm G. (ed), Akten des IV. Internationalen Kongresses für Hethitologie, Würzburg, 4.-8. Oktober 1999: pp. 532-540. Harrassowitz, Wiesbaden.

98 Water monuments in Hittite and Neo-Hittite periods: structure, functions, and connection with the “other world” Payne A., 2019, Hieroglyphic Inscriptions: Archaeologies of Sacred Space, in Keil W.E. et al. (eds), Zeichentragende Artefakte Im Sakralen Raum: Zwischen Präsenz Und Unsichtbarkeit: pp. 245-262. De Gruyter, Berlin – New York. Rojas F., Sergueenkova, V., 2014, Traces of Tarhuntas: Greek, Roman, and Byzantine Interaction with Hittite Monuments, Journal of Mediterranean Archaeology 27/2: pp. 135-160. Equinox, Sheffield. Rossner E.P., 1988, Die hethitischen Felsreliefs in der Türkei. Eine archäologischer Reiseführer. 2. erweiterte Auflage. Eberhard P. Rossner Verlag, München, 252 pages. Şahin M., 1999, Neue Beobachtungen zum Felsrelief von Ivriz/Konya. Nicht in den Krieg, sondern zur Ernte: der Gott mit der Sichel, Anatolian Studies 49: pp. 165-176. British Institute at Ankara, London. Vidal-Naquet P., 1964, Histoire et idéologie: Karl Wittfogel et le concept de « mode de production asiatique », Annales 19: pp. 531-549. Armand Colin, Paris. Wittfogel K.A., 1957, Oriental Despotism: A Comparative Study of Total Power. Yale University Press, New Haven, 556 pages.

99 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Associazione Treviso Sotterranea (Treviso, Italy) 2   Gruppo Speleologico San Marco (Venezia, Italy) 3   Gruppo Naturalistico Montelliano (Nervesa della Battaglia, Treviso, Italy) *  Reference author: Massimiliano Zago, mobile: +39 3401134331 - [email protected] The ancient aqueducts of Asolo (Italy): new investigations and acquisitions Massimiliano Zago¹,*, Daniele Davolio M.², Marcello Pellegrini³, Roberto Sordi³, Marco Sordi³ Abstract Of the rare ancient works present in the Veneto region, the city of Asolo, the ancient Roman Acelum, preserves in its historic center two underground sections of Roman and medieval aqueducts, which are the best preserved in north-eastern Italy. These hypogea are known as the La Bot and Il Gattolo aqueducts, dating back to the 1st and 11th century AD, respectively. The uniqueness of the Roman aqueduct La Bot does not lie in its lenght, limited to 230 m , but in the four different coating techniques of the main specus, adopted for copying with the different lithological nature of the rocks crossed during its excavation. The Gattolo aqueduct, on the other hand, is a part of the city’s hydraulic infrastructure since the Middle Age, when the pristine hydraulic work, fallen into disuse, was replaced. The La Bot aqueduct has regained its original importance and function again since the Renaissance, thanks to the restoration work carried out by the Serenissima Republic of Venice. This contribution reconstructs the history of these two underground works, describing their architectural characteristics through a speleological interpretation that does not always coincide with that proposed by archaeologists in the recent past. Thanks to the recent acquisitions carried out with the use of new 3D topographic digital tools, the results of the latest researches are presented using innovative techniques, including an unprecedented 3D print model of the Roman specus. Keywords: specus, ancient aqueducts, Roman hydraulics, undergrounds ducts, Veneto, Italy. Historical-geographical framework The city of Asolo (province of Treviso) is located at an altitude of 210 m asl, on the hills of the pre-Alpine belt of the Veneto region, which here rises in two parallel ridges, with a maximum height of 380 m asl. The area of the historic center of the city is characterized by terrains made of variably cemented conglomerates interbedded, on average every 3 meters, by thinner layers of moderately cemented sands and thicker layers of clayey silt (Mondin, 2002). From a historical point of view, the first archaeological findings date the protohistoric settlements back to around the 10th century BC, but it is during the Roman imperial era that the city grew and acquired significant importance. The area of the adjacent high Venetian plain was interested, since the 1st century BC, by a profound agro-territorial organization known as “centuriation”, which involved the city of Acelum (Asolo) as an economic and administrative centre, already recognized at the time as a Roman Municipium. The city was a crossing for the huge commercial traffic of wool, minerals and timber between the Alpine area and the important cities of Padua (Padova) and Vicetia (Vicenza). Therefore Acelum was served by the nearby important Roman road Via Postumia, and crossed by the Via Aurelia that directly connected Padua to the northernmost Via Claudia Augusta towards the north, up to the current German Bavaria (Riera, 2006). Archaeological findings from the Roman imperial era, preserved today in the city’s Civic Museum, testify the presence of important city infrastructures, including a large Roman villa, a thermal establishment and a theatre. It will be precisely the thermal establishment, which stood in the area of the current square Brugnoli, which dictated the construction of hydraulic works for feeding the city with water coming from two nearby perennial springs, placed beyond an orographic obstacle located just north of the town. The Roman aqueduct La Bot The function of La Bot aqueduct was to convey the water of two springs located on the northern slopes of the Monte Ricco hill to the city. From the Tintina spring, located 450 m east of the city centre, water flowed inside fistulae (lead pipes) and then probably poured into a loading tank (castellum) at the entrance of the underground section of the aqueduct, mixing with the water from the nearby Franca source (Riera, 1991). The crossing of the relief located north of the city forced the